PATRICIA A.E. MIROTH; STANLEY R. MIROTH, Plaintiffs-Appellants, v. COUNTY OF TRINITY; TRINITY COUNTY HEALTH AND HUMAN SERVICES; TRINITY COUNTY CHILD WELFARE SERVICES; LIZ HAMILTON; MARIO ANGELONE; NICOLE HAYES BRADFORD; ALLISON BALLARD; MEGAN SHOLTY-SCALZO; ANGELA BERGLUND; ASHLEY POQUETTE; VERLIN JOHNSON; DEBBIE CARTER, Defendants-Appellees.
No. 23-15759
United States Court of Appeals, Ninth Circuit
FOR PUBLICATION
UNITED STATES COURT OF APPEALS
FOR THE NINTH CIRCUIT
PATRICIA A.E. MIROTH;
STANLEY R. MIROTH,
Plaintiffs-Appellants,
v.
COUNTY OF TRINITY; TRINITY
COUNTY HEALTH AND HUMAN
SERVICES; TRINITY COUNTY
CHILD WELFARE SERVICES; LIZ
HAMILTON; MARIO ANGELONE;
NICOLE HAYES BRADFORD;
ALLISON BALLARD; MEGAN
SHOLTY-SCALZO; ANGELA
BERGLUND; ASHLEY POQUETTE;
VERLIN JOHNSON; DEBBIE
CARTER,
Defendants-Appellees.
No. 23-15759
D.C. No.
2:22-cv-00460-
KJM-JDP
OPINION
for the Eastern District of California
Kimberly J. Mueller, Chief District Judge, Presiding
Argued and Submitted August 19, 2024
San Francisco, California
2 MIROTH V. COUNTY OF TRINITY
Filed May 8, 2025
Before: Marsha S. Berzon, Daniel A. Bress, and Lawrence
VanDyke, Circuit Judges.
Opinion by Judge Bress;
Dissent by Judge VanDyke
SUMMARY*
Rooker-Feldman Doctrine
The panel reversed the district court’s dismissal, for lack
of subject matter jurisdiction under the Rooker-Feldman
doctrine, of a federal civil rights lawsuit alleging that county
officials failed to provide social services to plaintiffs and
committed fraud in state child custody proceedings.
After concluding that plaintiffs’ children were at risk in
the family home, defendants obtained warrants to take the
children into protective custody. A California state court
subsequently terminated plaintiffs’ parental
rights. Following unsuccessful appeals in state court,
plaintiffs filed this lawsuit in federal court, alleging that
defendants (1) failed to provide plaintiffs with social
services required by state law, and (2) made false and
misleading statements to the state court resulting in the
termination of their parental rights. The district court
dismissed the federal claims under the Rooker-Feldman
been prepared by court staff for the convenience of the reader.
MIROTH V. COUNTY OF TRINITY 3
doctrine, finding that plaintiffs were seeking relief from the
state court judgments, and declined to exercise jurisdiction
over the state law claims.
The panel held that although plaintiffs’ lawsuit may
suffer from other infirmities, most notably preclusion, the
Rooker-Feldman doctrine did not deprive the district court
of subject matter jurisdiction over plaintiffs’ claims. The
Rooker-Feldman applies only when the federal plaintiff both
asserts as her injury legal error or errors by the state court
and seeks as her remedy relief from the state court
judgment. Here, plaintiffs’ claims in their operative second
complaint did not seek relief from or reversal of the state
court’s order. Rather, they sought money damages,
asserting that defendants denied them services and made
false claims to the court. Because these were legal wrongs
by adverse parties that preceded the issuance of the state
court order, the Rooker-Feldman doctrine did not apply.
Dissenting, Judge VanDyke agreed with the majority’s
statement of the law on the Rooker-Feldman doctrine, but
disagreed with its application of that law to the unique
circumstances in this case. Plaintiffs functionally sought a
remedy for an injury directly caused by a state court
judgment, placing this case squarely within Rooker-Feldman’s ambit. The claims in this case were all based on
intrinsic fraud in the state court proceedings, where plaintiffs
had an opportunity to present their claims and to rebut the
supposed fraud, not extrinsic fraud that would have kept the
state court from fully hearing plaintiffs’ claims. When a
litigant’s asserted injury flows from the state court’s decision
and the litigant expressly disclaims, as in this case, that
extrinsic fraud influenced that decision, then
Rooker-Feldman applies. The majority’s overly narrow
articulation turns Rooker-Feldman into a mere pleading
4 MIROTH V. COUNTY OF TRINITY
requirement that can be easily circumvented. Judge
VanDyke would affirm the district court.
COUNSEL
Shannon C. Wilhite (argued), Shannon C. Wilhite Attorney
at Law, Bayside, California, for Plaintiffs-Appellants.
John A. Whitesides (argued) and Serena M. Warner, Angelo
Kilday & Kilduff, Sacramento, California; Debbie Carter,
Pro Se, Susanville, California; for Defendants-Appellees.
OPINION
BRESS, Circuit Judge:
Under the Rooker-Feldman doctrine, see Rooker v.
Fidelity Trust Co., 263 U.S. 413 (1923); D.C. Ct. of Appeals
v. Feldman, 460 U.S. 462 (1983), federal district courts lack
subject matter jurisdiction over “cases brought by state-court
losers complaining of injuries caused by state-court
judgments rendered before the district court proceedings
commenced and inviting district court review and rejection
of those judgments.” Exxon Mobil Corp. v. Saudi Basic
Indus. Corp., 544 U.S. 280, 284 (2005). We consider this
often-misunderstood doctrine in the context of a federal civil
rights lawsuit alleging that county officials failed to provide
social services and committed fraud in state child custody
proceedings. Although the plaintiffs’ lawsuit may suffer
from other infirmities, most notably preclusion, the Rooker-
MIROTH V. COUNTY OF TRINITY 5
Feldman doctrine did not deprive the district court of
jurisdiction.
We reverse and remand for further proceedings.
I
We recite the allegations of plaintiffs’ operative second
amended complaint. Plaintiffs Patricia and Stanley Miroth
are the parents of minor children A.M. and S.M. After
concluding that the children were at risk in the family home,
officials in Trinity County, California obtained warrants to
take the children into protective custody. Later, a California
state court terminated the Miroths’ parental rights. At the
time they were taken into state custody, A.M. was five years
old and S.M. was five days old.
The events that led to the Miroths losing custody of their
children date back to February 2018, when the County’s
Child Welfare Services (CWS) received a report that Lonnie
Smith, a registered sex offender and Patricia Miroth’s
stepfather, was living in the home. CWS employee Megan
Sholty-Scalzo visited the Miroths’ home and confirmed the
report. No action was immediately taken; the Miroths allege
that they were not told to remove Smith from the home. In
spring 2018, CWS received reports that the Miroths were
engaging in physical and verbal altercations with each other.
The Miroths contend that Sholty-Scalzo and other
County officials were aware of the difficulties in the Miroth
home but
plan. Such a plan would have established a pathway for the
Miroths to improve their home environment and maintain
custody of the children. Between March 29, 2018, and April
30, 2018, Patricia Miroth left numerous phone messages and
emails for Sholty-Scalzo seeking assistance. Sholty-Scalzo
6 MIROTH V. COUNTY OF TRINITY
allegedly did not return any of the messages. According to
the second amended complaint, Sholty-Scalzo and other
County officials improperly concluded, based on Patricia
Miroth’s past experiences involving her other children, that
Patricia was incapable of improving as a parent, when a
safety plan would have instead allowed the Miroths to
establish goals and expectations that would help them
maintain custody of A.M. and S.M.
In May 2018, CWS obtained a protective custody
warrant to take custody of A.M., the only child in the house
at the time. A warrant was similarly issued for S.M.
following his birth in August 2018. In March 2019, the state
court permanently terminated the Miroths’ parental rights to
A.M. and S.M. The Miroths’ second amended complaint
alleges that County defendants procured this result by
defrauding the state court about Patricia Miroth’s fitness as
a parent and about whether the County had provided the
Miroths with required social services. In particular, the
Miroths allege that County officials before the state court
(1) falsely stated that they provided services, when they
failed to develop a safety plan; (2) misrepresented the facts
surrounding the earlier termination of Patricia Miroth’s
parental rights for her other children; (3) falsely stated that
Patricia Miroth used drugs; (4) withheld exculpatory
evidence; and (5) lied to the court about whether the Miroths
understood the nature of their actions, learned from their past
mistakes, or made progress toward improving the
atmosphere within the home.
Following unsuccessful appeals in state court, the
Miroths in March 2022 filed this lawsuit in federal court
against, inter alia, the County of Trinity, Sholty-Scalzo, and
various County employees. The Miroths asserted various
claims under
MIROTH V. COUNTY OF TRINITY 7
request for relief, the Miroths specifically asked the district
court to “[r]everse the termination of TRISH and STAN’s
parental rights, and make orders to reunify children and
parents.” The Miroths subsequently filed a first amended
complaint that similarly sought to void the decision of the
state court, arguing that the “only just remedy [was] to
reinstate their parental rights as to A.M. and S.M.” Citing
these explicit requests to overturn the state court judgment,
the district court dismissed the Miroths’ first amended
complaint under the Rooker-Feldman doctrine, with leave to
amend.
The Miroths then filed the operative second amended
complaint. Unlike their two prior complaints, the Miroths
this time dropped any request to reinstate their parental
rights. Instead, they sought only money damages under
The district court again dismissed the federal claims
under the Rooker-Feldman doctrine and declined to exercise
jurisdiction over the state law claims. The court held that
because the Miroths “still essentially ask the court to review
the rulings of the state court and find they were in error,” the
lawsuit remained “a forbidden de facto appeal” of a state
court judgment. Although the Miroths were no longer
asking to have their parental rights reinstated, they were
nonetheless “assert[ing] the alleged errors of the state court
as their legal injury.” Because the Miroths had dropped
certain allegations concerning “extrinsic fraud,” the district
court viewed
challenging the state court decisions.”
After taking judicial notice of the records of the state
court proceedings, the district court concluded that “all of
the alleged fraud, misrepresentation[,] and omissions—
8 MIROTH V. COUNTY OF TRINITY
regarding the provision of services, drug use, counseling,
removal of children over fourteen years ago, progress made
by plaintiffs—were matters brought to the state court’s
attention . . . or were matters within plaintiffs’ knowledge.”
The Miroths were also seeking “relief from the state court
judgments” within the meaning of Rooker-Feldman because
their federal lawsuit “seek[s] relief from the consequences of
those judgments.”
The district court dismissed the Miroths’ second
amended complaint without leave to amend. This appeal
followed. We review the district court’s dismissal for lack
of jurisdiction under Rooker-Feldman de novo. Noel v. Hall,
341 F.3d 1148, 1154 (9th Cir. 2003).
II
A
With very limited exceptions, the United States Supreme
Court is the only federal court empowered to review the final
judgments of state courts.
Mobil, 544 U.S. at 283; Noel, 341 F.3d at 1154–55. As a
negative inference from this statutory grant of authority to
the Supreme Court, federal district courts may not entertain
appeals of state court judgments. Exxon Mobil, 544 U.S. at
283; Noel, 341 F.3d at 1154–55; Kougasian v. TMSL, Inc.,
359 F.3d 1136, 1139 (9th Cir. 2004). An appeal of a state
court judgment filed in federal district court—essentially an
appeal filed in the wrong court—is a “forbidden de facto
appeal” over which a district court lacks subject matter
jurisdiction. Kougasian, 359 F.3d at 1139. This deceptively
simple proposition, known as the Rooker-Feldman doctrine,
has led to a good deal of misunderstanding over the years,
with lower federal courts struggling to evaluate their
MIROTH V. COUNTY OF TRINITY 9
jurisdiction in cases involving parties who had previously
litigated against each other in state court.
The Supreme Court has found that Rooker-Feldman
precluded jurisdiction only twice, in the doctrine’s
eponymous cases, Rooker v. Fidelity Trust Co., 263 U.S. 413
(1923), and District of Columbia Court of Appeals v.
Feldman, 460 U.S. 462 (1983). We begin our analysis with
these cases, which demonstrate the limited scope of Rooker-Feldman’s restraint on the subject matter jurisdiction of
federal district courts.
In Rooker, a litigant sought to have a federal district
court declare an Indiana state court decision “null and void”
based on the state court’s legal errors in deciding the case.
263 U.S. at 414–15. The Supreme Court held that the
dispute was “plainly not within the District Court’s
jurisdiction as defined by Congress.” Id. at 415. Had the
district court exercised jurisdiction, it would have been “an
exercise of appellate jurisdiction,” when “[t]he jurisdiction
possessed by the District Courts is strictly original.” Id. at
416. “Under the legislation of Congress, no court of the
United States other than [the Supreme Court] could entertain
a proceeding to reverse or modify the judgment for errors of
that character.” Id. at 416. Thus, the district court lacked subject
matter jurisdiction over the suit.
In Feldman, decided sixty years later, Marc Feldman and
Edward Hickey sought to practice law in the District of
Columbia. Because the two had not studied at a law school
accredited by the American Bar
Columbia Court of Appeals denied them admission under a
local court rule. 460 U.S. at 467–68, 470–72. Feldman and
Hickey sought to have a federal district court overturn the
D.C. Court of Appeals’ decision. Id. at 468–69, 472–73.
10 MIROTH V. COUNTY OF TRINITY
But they also asked the federal court to find that the D.C.
Court of Appeals had violated their due process rights by
“act[ing] arbitrarily and capriciously in denying their
petitions for [a] waiver” of the requirement that they study
at an accredited school. Id. at 486; see also id. at 469 n.3.
The Supreme Court noted that Congress had in 1970
made “final judgments and decrees of the District of
Columbia Court of Appeals . . . reviewable by the Supreme
Court of the United States” on the same basis as decisions of
state supreme courts. Id. at 464 (quoting 84 Stat. 475, D.C.
Code § 11–102). When Congress had directed that review
of an appeal from the D.C. Court of Appeals would take
place in the Supreme Court, it divested lower courts of
jurisdiction over these appeals, just as Rooker held that
federal district courts lack jurisdiction over appeals from
state courts. Id. at 464. Recognizing that Feldman and Hickey’s
first request was to overturn the D.C. Court of Appeals’
determination, the Supreme Court held that they “should
have sought review” in the Supreme Court, as “the District
Court lacked subject matter jurisdiction over their
complaints.” Id. at 482.
Although the allegation that the D.C. Court of Appeals
acted “arbitrarily and capriciously” was not a direct
challenge to its judgment, the Supreme Court found that the
district court also lacked subject matter jurisdiction to hear
that claim. Id. at 486–87. The Supreme Court
acknowledged that a “general challenge to the
constitutionality” of the D.C. court’s policy denying
admission to applicants who had not graduated from an
accredited law school would be within the district court’s
jurisdiction, as that would “not require review of a judicial
decision in a particular case.” Id. at 483, 487. But
determining whether the D.C. court’s decision was arbitrary
MIROTH V. COUNTY OF TRINITY 11
and capricious was “inextricably intertwined” with the D.C.
court’s decision to exclude Feldman and Hickey from the
bar, rendering it also outside of the district court’s
jurisdiction. Id. at 486–87.
Feldman’s use of the phrase “inextricably intertwined”
did not expand the universe of cases in which Rooker-Feldman deprives federal district courts of subject matter
jurisdiction; it only sets the bounds of that deprivation within
the context of a lawsuit in which the Rooker-Feldman
doctrine already applies. Our decision in Noel explained
this:
The premise for the operation of the
“inextricably intertwined” test in Feldman is
that the federal plaintiff is seeking to bring a
forbidden de facto appeal. The federal suit is
not a forbidden de facto appeal because it is
“inextricably intertwined” with something.
Rather, it is simply a forbidden de facto
appeal. Only when there is already a
forbidden de facto appeal in federal court
does the “inextricably intertwined” test come
into play: Once a federal plaintiff seeks to
bring a forbidden de facto appeal, as
in Feldman, that federal plaintiff may not
seek to litigate an issue that is “inextricably
intertwined” with the state court judicial
decision from which the forbidden de facto
appeal is brought.
341 F.3d at 1158; see also Bell v. City of Boise, 709 F.3d
890, 897 (9th Cir. 2013) (“The ‘inextricably intertwined’
language from Feldman is not a test to determine whether a
12 MIROTH V. COUNTY OF TRINITY
claim is a de facto appeal, but is
step in the Rooker-Feldman analysis. Should the action not
contain a forbidden de facto appeal, the Rooker-Feldman
inquiry ends.”).
B
In the decades following Feldman, circuit courts
diverged on the proper scope of the Rooker-Feldman
doctrine. Our court took a circumscribed view in Noel, our
seminal decision in this area. Noel held that “where the
federal plaintiff does not complain of a legal injury caused
by a state court judgment, but rather of a legal injury caused
by an adverse party, Rooker-Feldman does not bar
jurisdiction.” Noel, 341 F.3d at 1163. Under this
interpretation, “[i]t is a forbidden de facto appeal under
Rooker-Feldman when the plaintiff in federal district court
complains of a legal wrong allegedly committed by the state
court, and seeks relief from the judgment of that court.” Id.
Outside of this narrow situation, when a plaintiff complains
of harms caused by an adverse party in state court
proceedings, the limits on federal courts’ ability to grant
relief come from doctrines of abstention and comity (if the
federal plaintiff and adverse party are simultaneously
litigating in state court) or doctrines of preclusion (if the state
court suit has reached final judgment). Id. at 1163–64.
Unlike the Ninth Circuit, other circuits took an expansive
view of Rooker-Feldman as depriving federal district courts
of subject matter jurisdiction whenever the federal suit
would imply the invalidity of a state court judgment or when
it sought to litigate an issue that was or could have been
raised in state court. One circuit, for example, said that
“[t]he Rooker-Feldman doctrine is broad enough to bar all
federal claims which were, or should have been, central to
MIROTH V. COUNTY OF TRINITY 13
the state court decision, even if those claims seek a form of
relief that might not have been available from the state
court.” Goodman ex rel. Goodman v. Sipos, 259 F.3d 1327,
1333 (11th Cir. 2001); see also, e.g., FOCUS v. Allegheny
Cnty. Ct. of Common Pleas, 75 F.3d 834, 840 (3d Cir. 1996)
(“If the relief requested in the federal action requires
determining that the state court decision is wrong or would
void the state court’s ruling, then . . . the district court has no
subject matter jurisdiction to hear the suit.” (quoting
Charchenko v. City of Stillwater, 47 F.3d 981, 983 (8th Cir.
1995))); Moccio v. N.Y. State Off. of Ct. Admin., 95 F.3d 195,
199 (2d Cir. 1996) (approvingly citing sources maintaining
that “there is full overlap between the Rooker-Feldman doctrine and preclusion,” and positing that Rooker-Feldman could “reach[] beyond those situations in which the
federal action is precluded by res judicata or collateral
estoppel”).
In 2005, the Supreme Court put this debate to rest in
Exxon Mobil Corporation v. Saudi Basic Industries
Corporation, 544 U.S. 280, clarifying that Rooker-Feldman
occupies only the “narrow ground” exemplified in the
Rooker and Feldman cases themselves. Id. at 284. Exxon
Mobil explained that some lower courts had construed
Rooker-Feldman “to extend far beyond the contours of the
Rooker and Feldman cases, overriding Congress’s conferral
of federal-court jurisdiction concurrent with jurisdiction
exercised by state courts, and superseding the ordinary
application of preclusion law pursuant to
Id. at 283. The Supreme Court rejected this broader take on
Rooker-Feldman. And in doing so, its analysis tracked our
reasoning in Noel—which Exxon Mobil approvingly cited.
See id. at 293.
14 MIROTH V. COUNTY OF TRINITY
Exxon Mobil held that the Rooker-Feldman doctrine “is
confined to cases of the kind from which the doctrine
acquired
complaining of injuries caused by state-court judgments
rendered before the district court proceedings commenced
and inviting district court review and rejection of those
judgments.” Id. at 284. The existence of a prior state court
judgment and a federal plaintiff seeking to relitigate a matter
already litigated in state court were not circumstances
sufficient to invoke Rooker-Feldman, Exxon Mobil
explained, even if the federal plaintiff’s claim “denie[d] a
legal conclusion that a state court has reached in a case to
which he was a party.” Id. at 293 (quoting GASH Assocs. v.
Vill. of Rosemont, 995 F.2d 726, 728 (7th Cir. 1993)). In that
situation, “there is jurisdiction and state law determines
whether the defendant prevails under principles of
preclusion.” Id. at 293 (quoting GASH Assocs., 995 F.2d at 728).
In the years since Exxon Mobil, and consistent with our
earlier decision in Noel, we have continued to recognize the
limited scope of Rooker-Feldman’s restriction on the subject
matter jurisdiction of federal district courts. See, e.g., Cogan
v. Trabucco, 114 F.4th 1054, 1064 (9th Cir. 2024) (“Our
caselaw has further narrowed the doctrine as applying only
to suits alleging errors by the state courts in rendering
judgment, as opposed to misconduct by litigants in obtaining
such a judgment.”); Bell, 709 F.3d at 897 (“The court erred
by dismissing Plaintiffs’ claims for retrospective relief under
the Rooker-Feldman doctrine. Although Plaintiffs sought
relief designed to remedy injuries suffered from a state court
judgment, they did not allege before the court that the state
court committed legal error, nor did they seek relief from the
state court judgment itself.”). Post-Exxon Mobil, the
Supreme Court has likewise reiterated that the “cases since
MIROTH V. COUNTY OF TRINITY 15
Feldman have tended to emphasize the narrowness of the
Rooker-Feldman rule.” Lance v. Dennis, 546 U.S. 459, 464
(2006) (per curiam).
Although application of Rooker-Feldman and preclusion
doctrines can both lead to plaintiffs losing in federal court in
the aftermath of a state court judgment, Exxon Mobil was
clear that “Rooker-Feldman does not otherwise override or
supplant preclusion doctrine.” 544 U.S. at 284. We have
thus emphasized that Rooker-Feldman and preclusion “are
analytically distinct.” Cogan, 114 F.4th at 1064. Rooker-Feldman is a jurisdictional doctrine, whereas preclusion is
not. Exxon Mobil, 544 U.S. at 293. Improperly dismissing
for lack of subject matter jurisdiction under Rooker-Feldman
is therefore material error, for federal courts have a “virtually
unflagging obligation . . . to exercise the jurisdiction given
[to] them.” Colorado River Water Conservation Dist. v.
United States, 424 U.S. 800, 817 (1976). Further, under
is a matter of state law, Lance, 546 U.S. at 466, whereas the
reach of Rooker-Feldman is a question of federal law,
Cogan, 114 F.4th at 1064. Equating Rooker-Feldman with
preclusion “risks turning th[e] limited [Rooker-Feldman]
doctrine into a uniform federal rule governing the preclusive
effect of state-court judgments,” contrary to
In short, the Supreme Court’s cases from Rooker and
Feldman through Exxon Mobil and Lance, and our cases
from Noel onward, confirm that “Rooker-Feldman is not
simply preclusion by another name.” Lance, 546 U.S. at 466.
Concluding that a plaintiff’s claims are barred by issue
or claim preclusion
matter jurisdiction under Rooker-Feldman.
16 MIROTH V. COUNTY OF TRINITY
C
Special difficulties can arise under Rooker-Feldman
when the federal plaintiff sues an adverse party from a state
court proceeding and claims that the adverse party
fraudulently procured the state court judgment. Our cases
make clear that while principles of preclusion may
ultimately render the federal suit meritless, suits like this
may be maintained without running afoul of Rooker-Feldman. Two of our cases are most relevant here:
Kougasian v. TMSL, Inc., 359 F.3d 1136 (9th Cir. 2004), and
Benavidez v. County of San Diego, 993 F.3d 1134 (9th Cir.
In Kougasian, the plaintiff, Dawn Kougasian, lost two
wrongful death suits in California state court and then filed
suit in federal court. 359 F.3d at 1138. In her federal
complaint, she alleged, inter alia, that the defendants had
filed a false declaration on which the state court dispositively
relied, and that “defendants prevented her from challenging
the declaration by presenting it to the court at the last minute
and by refusing to provide the declarant’s telephone number
or address.” Id. In one of her causes of action, Kougasian
alleged that the defendants had obtained their judgments in
state court “through extrinsic fraud on the court.” Id. at
1139. Kougasian did not allege legal error by the state courts
in rendering their judgments, but instead sought to “set aside
these judgments based on the alleged extrinsic fraud by
defendants that produced those judgments.” Id. at 1139.
We held that Kougasian’s fraud-based claims did not
violate the Rooker-Feldman doctrine. We explained that
under California law, “extrinsic fraud”—“conduct which
prevents a party from presenting his claim in court”—is a
basis for setting aside a prior judgment. Id. at 1140 (quoting
MIROTH V. COUNTY OF TRINITY 17
Wood v. McEwen, 644 F.2d 797, 801 (9th Cir. 1981)). Even
though in this situation “[i]t is clear” that “the plaintiff is
seeking to set aside a state court judgment,” Rooker-Feldman still was not implicated because “[e]xtrinsic fraud
on a court is, by definition, not an error by that court. It is,
rather, a wrongful act committed by the party or parties who
engaged in the fraud.” Id. at 1140–41. Rooker-Feldman
therefore did not apply because Kougasian was alleging “a
wrongful act by the adverse party” and not “legal error by
the state court.” Id. at 1141; see also Wallingford v. Bonta,
82 F.4th 797, 819 (9th Cir. 2023) (Collins, J., dissenting)
(“Although the plaintiffs’ federal suit in Kougasian
unquestionably did seek relief from the prior adverse state
court judgments, we held that the [Rooker-Feldman]
doctrine nonetheless did not apply because the plaintiffs
were not ‘alleging a legal error by the state court as the basis
for that relief.’” (quoting Kougasian, 359 F.3d at 1140)).1
In Benavidez, which we decided after Exxon Mobil, a
state juvenile court issued orders authorizing a medical
examination of two children whom social workers had
removed from their parents’ home. 993 F.3d at 1140. As a
result of the orders, county officials conducted a “full body
inspection” of the children, performed blood and urine tests,
and administered vaccinations. Id. at 1141. The parents
filed suit in federal court under
that the inspection of the children, without notice to or
consent of the parents, violated the parents’ constitutional
workers had engaged in “judicial deception” by making
grounds. See 82 F.4th at 799.
18 MIROTH V. COUNTY OF TRINITY
misrepresentations to the juvenile court, which caused the
court to order the examinations. Id. at 1144.
Although the medical examinations took place pursuant
to an order of a state court, we held that Rooker-Feldman did
not apply. Id. at 1143. “Despite the judicial context and
intermediate step of the juvenile court Orders,” we
explained, the parents’ “claims d[id] not seek relief from or
reversal of the juvenile court’s Orders.” Id. The parents’
claims were instead based on “a legal wrong by [the social
workers] preceding the issuance of the Orders.” Id. And
these alleged wrongs could not “avoid scrutiny” under
Rooker-Feldman merely “because they were successful in
deceiving the juvenile court.” Id. Indeed, we went on,
“[e]ven if the [parents] had directly challenged the juvenile
court decision, which they did not, the extrinsic fraud
corollary to the Rooker-Feldman doctrine would apply”
under Kougasian. Id. Because the parents did not learn of
the medical examinations until later, they could have gone
so far as to “directly challenge” the state court decision,
without running afoul of Rooker-Feldman. Id. at 1144.
III
We now turn to whether the Rooker-Feldman doctrine
deprived the district court of subject matter jurisdiction over
the Miroths’ claims. We hold it did not.
A
As we discussed above, the Miroths’ second amended
complaint can be broadly categorized into two types of
allegations: that county officials (1) failed to provide the
Miroths with social services required by state law, and
(2) made false and misleading statements to the state court
that caused the court to terminate the Miroths’ parental rights
MIROTH V. COUNTY OF TRINITY 19
over A.M. and S.M. Neither allegation implicates the
Rooker-Feldman doctrine. Although that doctrine has
developed a reputation for complexity, the reasons it does
not apply here are straightforward.
Rooker-Feldman “applies only when the federal plaintiff
both asserts as her injury legal error or errors by the state
court and seeks as her remedy relief from the state court
judgment.” Kougasian, 359 F.3d at 1140. Neither element
is met here. As an initial matter, as in Benavidez, “[d]espite
the judicial context and intermediate step of the [state] court
Orders, the [Miroths’] claims do not seek relief from or
reversal of the [state] court’s Orders. Therefore, the Rooker-Feldman doctrine does not serve as a jurisdictional bar to
their claims here.” 993 F.3d at 1143. Although the Miroths
in prior iterations of their complaint did seek to set aside the
state court judgment and reinstate their parental rights, their
operative second complaint dropped these requests for relief.
This is a sufficient basis by which to conclude that the
Rooker-Feldman doctrine does not apply.
But beyond this, and perhaps more fundamentally, the
Miroths’ second amended complaint does not contravene
Rooker-Feldman because it “does not assert ‘as a legal
wrong an allegedly erroneous decision by a state court,’ but
rather ‘an allegedly illegal act or omission by an adverse
party.’” Kougasian, 359 F.3d at 1140 (quoting Noel, 341
F.3d at 1164). The Miroths’ allegation that the defendants
failed to develop a required safety plan to ensure that the
Miroths could retain custody of the children asserts a wrong
by adverse parties, not the state court. The same is true of
the Miroths’ allegations that the
state court into terminating the Miroths’ parental rights. The
second amended complaint alleges that the defendants
“misrepresented” facts to the state court, “concealed known
20 MIROTH V. COUNTY OF TRINITY
facts from the Court,” and “falsely represented facts, both
written and oral.” These are legal wrongs by adverse parties
“preceding the issuance of” the state court orders.
Benavidez, 993 F.3d at 1143. These “alleged legal wrongs
by [adverse parties] cannot avoid scrutiny” under Rooker-Feldman merely because, by the allegations of the second
amended complaint, “they were successful in deceiving the
[state] court.” Id. (quotations omitted). As we recently
reiterated, “[o]ur caselaw has further narrowed the doctrine
as applying only to suits alleging errors by the state courts in
rendering judgment, as opposed to misconduct by litigants
in obtaining such a judgment.” Cogan, 114 F.4th at 1064.
This principle governs here.2
Thus, Rooker-Feldman does not apply, and the district
court had subject matter jurisdiction over the case.
B
The district court gave several reasons for finding that
the Rooker-Feldman doctrine barred the Miroths’ claims.
The dissent echoes some of those grounds. Because of the
confusion that Rooker-Feldman has engendered over the
years, we take this opportunity to explain why the district court’s reasoning was mistaken.
The district court first found that the Miroths “assert the alleged errors of the state court as their legal injury” because their second amended complaint dropped certain allegations of extrinsic fraud. According to the district court, this “withdrawal of the extrinsic fraud argument” was “dispositive,” demonstrating that “plaintiffs are now directly challenging the state court decisions.” The dissent endorses this theory as well.
This reasoning rests on a misunderstanding of both the Miroths’ allegations and the relevance of formal allegations of extrinsic fraud to the Rooker-Feldman analysis. As to the former, the Miroths’ original and first amended complaint both included two paragraphs of allegations, more in the form of legal argument, concerning extrinsic fraud. These paragraphs asserted that “extrinsic fraud exists in this case” and “is an exception to the Rooker-Feldman doctrine,” and argued why Rooker-Feldman should not apply. The Miroths then removed these paragraphs from their second amended complaint. In the district court, the Miroths explained they had done so because once they were no longer “requesting relief in the form of a reversal of the state court decision,” they were outside of Rooker-Feldman for that reason alone, and they therefore believed it was no longer necessary to allege extrinsic fraud.
The Miroths’ removal of allegations (again, more legal argument) relating to extrinsic fraud did not create a Rooker-Feldman problem. The “extrinsic fraud corollary” to Rooker-Feldman comes into
Therefore, and although the Miroths could have perhaps telegraphed their pleading strategy in a more accessible way, the Miroths withdrawing the extrinsic fraud allegations did not mean they were thereby alleging an injury caused by the state court, based on the state court’s legal errors. The Miroths did not walk into a Rooker-Feldman sinkhole the moment they excised “extrinsic fraud” from their complaint. Instead, regardless of the legal labels attached to the allegations, the second amended complaint alleges that Sholty-Scalzo and other defendants defrauded the state court into issuing its order terminating the Miroths’ parental rights. Those allegations assert wrongful conduct by adverse parties in litigation, not by the state court itself. They thus do not implicate Rooker-Feldman. See Cogan, 114 F.4th at 1064; Benavidez, 993 F.3d at 1143; Kougasian, 359 F.3d at 1139–41; Noel, 341 F.3d at 1155–57. The dissent’s view that the Miroths’ asserted injuries “necessarily travel through the state court’s judgment” is inconsistent with Kougasian, Benavides, and, indeed, our entire Rooker-Feldman jurisprudence.
The district court’s analysis, and more so the dissent, also rest on a broader misunderstanding of the analytical relevance of extrinsic fraud allegations in the context of Rooker-Feldman. As we described it in Kougasian, extrinsic fraud is a specific type of fraud claim that, at least under California law, provides a basis for setting aside a prior judgment. See Kougasian, 359 F.3d at 1140 (citing Zamora v. Clayborn Contracting Grp., Inc., 47 P.3d 1056, 1063 (Cal. 2002)). And our point in Kougasian was that because “[e]xtrinsic fraud on a court is, by definition, not an error by that court” but, “rather, a wrongful act committed by the party or parties who engaged in the fraud,” it follows that Rooker-Feldman “does not bar subject matter jurisdiction when a federal plaintiff alleges a cause of action for extrinsic fraud on a state court.” Id. at 1141.
But to say that allegations of extrinsic fraud definitionally do not cross Rooker-Feldman does not mean the absence of such allegations triggers Rooker-Feldman, even for fraud-based claims. Contrary to the dissent, we have never distinguished between so-called “intrinsic” fraud and extrinsic fraud in the Rooker-Feldman context. Nor would it be proper or helpful to import that state-law distinction into the federal-law question of the scope of the Rooker-Feldman doctrine.
Fraud allegations against an adverse party in litigation can avoid Rooker-Feldman even when they are not couched in extrinsic fraud terms. That was partially the case in Kougasian, in which we recognized that to the extent the plaintiff’s fraud causes of action were not based in extrinsic fraud, they still did not countermand Rooker-Feldman because they were based on “alleged wrongful acts by the defendants.” Kougasian, 359 F.3d at 1141; see also id. at 1140 n.1 (explaining that “[t]o the extent that Kougasian’s two causes of action for fraud and for abuse of process are not based on extrinsic fraud, . . . [t]hey are nevertheless not barred by Rooker-Feldman, but for reasons given in the analysis in the next section of this opinion”).
The same was true in Benavidez. There, we first held that the Rooker-Feldman doctrine did not apply because the plaintiffs “challenge[d] a legal wrong” by two
These cases show, contrary to the dissent, that alleging extrinsic fraud is one way of avoiding Rooker-Feldman, but it is not somehow a requirement, even when one challenges an adverse party’s alleged misconduct in a prior state court litigation. Indeed, if anything, it is the cause of action for extrinsic fraud that created the closer Rooker-Feldman question, because (at least under California law) it provides the basis for setting aside a state court judgment. See Kougasian, 359 F.3d at 1140–41 (“At first glance, a federal suit alleging a cause of action for extrinsic fraud on a state court might appear to come within the Rooker–Feldman doctrine.”).
To be sure, a plaintiff who cannot show extrinsic fraud, i.e., that he was prevented from presenting his fraud claim in court, id. at 1140, may have a more difficult time avoiding preclusion obstacles. But that does not change the fact that a plaintiff alleging fraud by another party in litigation is not alleging a legal error by the state court. See Benavidez, 993 F.3d at 1143; Kougasian, 359 F.3d at 1139–41.
This brings us to the second of the district court’s main rationales for finding that Rooker-Feldman applied, which is that based on its review of the state court record, of which the district court took judicial notice, “the state court heard evidence regarding defendants’ alleged misrepresentation” and evidently found it unpersuasive, given that it terminated the Miroths’ parental rights. According to the district court, “all of the alleged fraud, misrepresentation[,] and omissions—regarding the provision of services, drug use, counseling, removal of children over fourteen years ago, progress made by plaintiffs—were matters brought to the state court’s attention . . . or were matters within plaintiffs’ knowledge.”
The problem with this reasoning is that it does not detract from or disprove the fact that plaintiffs are alleging “as a legal wrong an allegedly illegal act or omission by an adverse party,” for which “Rooker-Feldman does not bar jurisdiction.” Kougasian, 359 F.3d at 1140 (quoting Noel, 341 F.3d at 1164). A careful examination of the state court record may well lead to the conclusion that the Miroths’ claims are barred by doctrines of preclusion (an issue we do not reach). But as the Supreme Court has made clear, “Rooker-Feldman does not otherwise override or supplant preclusion doctrine,” Exxon Mobil, 544 U.S. at 284, nor is Rooker-Feldman “simply preclusion by another name,” Lance, 546 U.S. at 466. Even if the Miroths are raising matters that were raised or could have been raised in state court, that does not mean they are alleging legal errors by the state court or seeking review and rejection of the state court judgment itself. The former is the domain of preclusion doctrines; the latter is Rooker-Feldman, whose strictures are not met here.
Finally, the district court held, and the dissent agrees, that the Miroths “seek relief from the state court judgments” because through their damages requests, “plaintiffs are seeking relief from the consequence of those judgments.” But “seek[ing] relief from a state court judgment,” Noel, 341 F.3d at 1164, which can implicate Rooker-Feldman, is not the same thing as seeking relief that would ameliorate the effects of an adverse state court judgment. In past cases, plaintiffs sought
Here, moreover, the true “consequences” of the state court judgment is that the Miroths lost custody of their children. Nothing in the Miroths’ second amended complaint, however, asks a federal court to address those consequences and reinstate the Miroths’ parental rights or vacate or otherwise reverse the state court orders. It is these types of requests for relief that can present Rooker-Feldman problems. See, e.g., Exxon Mobil, 544 U.S. at 291–92 (noting that in both Rooker and Feldman, the plaintiffs “called upon the District Court to overturn an injurious state-court judgment”); Kougasian, 359 F.3d at 1140 (noting that Rooker-Feldman applies only when a federal plaintiff “seeks as her remedy relief from the state court judgment”); Noel, 341 F.3d at 1156 (noting that the doctrine applies when a plaintiff “seeks to vacate or set aside the judgment of” a state court). But that is not the relief the Miroths sought here.
C
Our dissenting colleague offers some additional reasons why Rooker-Feldman should apply in this case. In our respectful view, the dissent’s approach to Rooker-Feldman is well afield of the doctrine’s proper scope. Whereas the district court’s errors were premised in part on its misconstruing the Miroths’ allegations, the dissent misconstrues much about the Rooker-Feldman doctrine itself. Indeed, the dissent’s views, if accepted, would rewind the clock to the bad old days in which some courts (though not ours) improperly applied Rooker-Feldman whenever the federal suit would cast doubt on a state court decision. The dissent reflects an expansive vision of Rooker-Feldman that the Supreme Court and this court have already rejected.
First, the dissent claims that the distinction between injuries caused by a state court’s legal errors in issuing a judgment and an adverse party’s wrongs in procuring that judgment is a “mere semantic shift,” because these injuries “are the same thing.” But a central premise of the Rooker-Feldman doctrine, as properly understood, is that these are not the same thing. Our cases could not be clearer:
If a federal plaintiff asserts as a legal wrong an allegedly erroneous decision by a state court, and seeks relief from a state court judgment based on that decision, Rooker-Feldman bars subject matter jurisdiction in federal district court. If, on the other hand, a federal plaintiff asserts as a legal wrong an allegedly illegal act or omission by an adverse party, Rooker-Feldman does not bar jurisdiction.
The dissent protests that in this case, none of the defendants’ allegedly false statements “injured plaintiffs apart from the state court’s judgment.” According to the dissent, “Plaintiffs were injured only because the state court relied on those statements in ordering the removal of plaintiffs’ children.” But when plaintiffs come to federal court following the completion of state court proceedings and raise allegations that relate to the same subject
Second, the dissent emphasizes that the Miroths had “ample opportunity” in state court to raise the fraud allegations they raise here, and that “the state court considered whatever responses the plaintiffs provided and simply disagreed with them.” But the dissent would once again improperly fuse preclusion with Rooker-Feldman. The Supreme Court has clearly held that a district court does not lose subject matter jurisdiction “simply because a party attempts to litigate in federal court a matter previously litigated in state court.” Exxon Mobil, 544 U.S. at 293. Instead, “[i]f a federal plaintiff ‘present[s] some independent claim, albeit one that denies a legal conclusion that a state court has reached in a case to which he was a party . . . , then there is jurisdiction and state law determines whether the defendant prevails under principles of preclusion.’” Id. (quoting GASH Assocs., 995 F.2d at 728). That the Miroths’ lawsuit, if successful, might imply the invalidity of the state court’s judgment does not mean that Rooker-Feldman applies.
Third, the dissent modifies the legal standards to create the misimpression that Rooker-Feldman is more embracing than it really is. Seizing on our description of Rooker-Feldman as encompassing “forbidden de facto appeal[s]” of state court judgments, Noel, 341 F.3d at 1156, the dissent claims that “[t]he use of words like ‘de facto’ . . . would be unnecessary if Rooker-Feldman applied only to express requests to reverse a state court.” From this softening of Rooker-Feldman’s core premise, the dissent then asserts that the question under Rooker-Feldman is “whether the federal suit is effectively a request for the federal court to remedy harm caused by the state court’s decision.”
But the dissent’s Rooker-Feldman is not the one the Supreme Court has given us, which is “confined to” “cases brought by state-court losers complaining of injuries caused by state-court judgments rendered before the district court proceedings commenced and inviting district court review and rejection of those judgments.” Exxon Mobil, 544 U.S. at 284. Our use of the term “de facto appeal” does not provide support for the dissent’s self-described “functional” approach to Rooker-Feldman. “De facto appeal” does not mean a more flexible Rooker-Feldman than Exxon Mobil or our cases allow. It simply means that a plaintiff has filed in federal district court what should have been an appeal filed in a higher state court or the Supreme Court. “Forbidden de facto appeal” is the conclusion one reaches after applying the proper tests for Rooker-Feldman; it is not itself the test. The same is true of Kougasian’s use of the word “tantamount” (which we quoted from Barrow v. Hunton, 99 U.S. (9 Otto) 80, 82–83 (1878)). See Kougasian, 359 F.3d at 1141. The quotation of one word in one case did not usher in the dissent’s entirely different approach to the Rooker-Feldman doctrine.
The dissent again manipulates the legal standards in claiming that Rooker-Feldman applies because “plaintiffs’ relief ‘is contingent upon a finding that the state court decision was in error’” (quoting Cooper v. Ramos, 704 F.3d 772, 782 (9th Cir. 2012)). This language from Cooper is taken out of context. In Cooper, we held that the plaintiff’s first claim was “a pure horizontal appeal of the state court’s decision” that “attacks the Superior Court judgment explicitly.” 704 F.3d at 779–80. It therefore “simply cannot be said that, here, Cooper, ‘does not challenge the adverse state court decision itself.’” Id. at 780 (brackets omitted) (quoting Skinner v. Switzer, 562 U.S. 521, 532 (2011)). Rooker-Feldman therefore applied. We then held that the remaining claims were “inextricably intertwined” with this first claim. Id. at 781–83.
The “contingent upon” language on which the dissent relies came from Cooper’s application of the “inextricably intertwined” test. But as we discussed above, our cases are clear that “inextricably intertwined” “is not a test to determine whether a claim is a de facto appeal.” Bell, 709 F.3d at 897. Instead, “[o]ur circuit has emphasized that ‘[o]nly when there is already a forbidden de facto appeal in federal court does the ‘inextricably intertwined’ test come into play.” Cooper, 704 F.3d at 778 (quoting Noel, 341 F.3d at 1158). The dissent therefore again miscasts language from our cases to suggest that Rooker-Feldman’s expanse is broader than precedent directs.
Fourth, the dissent claims that our decision conflicts with decisions from other circuits. That charge is unfounded and once again exemplifies the dissent’s misunderstanding of the Rooker-Feldman doctrine. Some of the cases the dissent cites held that the Rooker-Feldman doctrine did not apply. See Great W. Mining & Min. Co. v. Fox Rothschild LLP, 615 F.3d 159, 173 (3d Cir. 2010); McCormick v. Braverman, 451 F.3d 382, 392–93 (6th Cir. 2006). And in other cases the dissent cites, courts held that Rooker-Feldman barred jurisdiction when the plaintiff sought to overturn the state court judgment itself. See, e.g., RLR Invs., LLC v. City of Pigeon Forge, 4 F.4th 380, 383 (6th Cir. 2021) (holding that Rooker-Feldman applied when the plaintiff “filed a complaint in federal court alleging that the [state court] Order was unconstitutional and inconsistent with Tennessee law, asking the federal court to enjoin the Order’s enforcement”); Klimowicz v. Deutsche Bank Nat’l Tr. Co., 907 F.3d 61, 66 (1st Cir. 2018) (holding that Rooker-Feldman applied because “the plaintiff’s federal suit sought to vacate and set aside the Land Court’s final judgment of foreclosure,” “which would include enjoining enforcement of the Housing Court’s Order”). These cases all support our decision in this case, as well as the recognized distinctions that we follow—distinctions the dissent elides.
The dissent’s heavy reliance on Hoblock v. Albany County Board of Elections, 422 F.3d 77 (2d Cir. 2005), likewise demonstrates its misperception of governing law. Invoking a hypothetical lawsuit of a parent challenging a state court child custody termination, the Second Circuit observed that “[i]f the father sues in federal court for the return of his son on grounds that the judgment violates his federal substantive due-process right as a parent, he is complaining of an injury caused by the state judgment and seeking its reversal.” Id. at 87. That is true, but what the Miroths are seeking here is different; they are not requesting the return of their children.
The Miroths are not challenging the actions of County officials that were “pursuant to” or “produced by” a state court judgment, i.e., their actions executing the judgment. The Miroths are instead alleging that the County fraudulently procured the judgment, which is conduct by adverse parties that preceded the judgment. And again, the cases are clear that such suits can be maintained without crossing Rooker-Feldman. It bears repeating: “[o]ur caselaw has further narrowed the doctrine as applying only to suits alleging errors by the state courts in rendering judgment, as opposed to misconduct by litigants in obtaining such a judgment.” Cogan, 114 F.4th at 1064. That was the nature of the claims in both Benavides and Kougasian, in which we held that the Rooker-Feldman doctrine did not apply. Benavidez, 993 F.3d at 1143; Kougasian, 359 F.3d at 1139–41. Indeed, that was the nature of the claims in cases the dissent cites, in which courts likewise held that the Rooker-Feldman doctrine did not bar jurisdiction. See McCormick, 451 F.3d at 392 (holding that Rooker-Feldman did not apply because “Plaintiff asserts independent claims that those state court judgments were produced by certain Defendants through fraud, misrepresentation, or other improper means”); Great W. Mining & Min. Co., 615 F.3d at 170–73 (same, citing like cases from the Seventh Circuit); see also, e.g., Behr v. Campbell, 8 F.4th 1206, 1213–14 (11th Cir. 2021) (holding that Rooker-Feldman did not bar claims alleging that adverse parties committed fraud in state child custody proceedings).
Finally, the dissent repeatedly expresses concern about our assertedly narrow application of Rooker-Feldman and the possibility that plaintiffs may craft pleadings to avoid Rooker-Feldman’s prohibition on de facto appeals. Neither concern is warranted. Although we of course agree with the dissent that there is no “express” language or “specific words requirement” for Rooker-Feldman to apply, narrowness is a hallmark of the Rooker-Feldman doctrine. The cases repeatedly drive this point home. See Exxon Mobil, 544 U.S. at 284 (referring to the “narrow ground occupied by Rooker-Feldman”); id. at 291 (making clear the doctrine applies only in “limited circumstances”); Lance, 546 U.S. at 464 (“[O]ur cases since Feldman have tended to emphasize the narrowness of the Rooker-Feldman rule.”); id. (“narrow doctrine”); Cooper, 704 F.3d at 778 (“very sparing” doctrine); Behr, 8 F.4th at 1212 (“In short, district courts should keep one thing in mind when Rooker-Feldman is raised: it will almost never apply.”). Indeed, the Supreme Court has found that the doctrine precluded jurisdiction only twice. See Exxon Mobil, 544 U.S. at 283 (citing Rooker and Feldman). The dissent’s concern with our application of Rooker-Feldman is simply a disagreement with the limited manner in which the doctrine operates.
Nor is there any reason to be concerned about the supposed ill of pleadings that are carefully drafted to avoid Rooker-Feldman. There is nothing new about plaintiffs drafting complaints to account for jurisdictional or other procedural impediments. And as our case law demonstrates, plaintiffs are entitled to bring challenges that do not run afoul of Rooker-Feldman. This is not a mere matter of pleading particular
In fact, perhaps aware that Rooker-Feldman presented a difficult basis for affirmance, the defendants led off their answering brief by asking us to affirm on the alternative ground that the Miroths’ claims are barred under principles of preclusion. As we noted above, many of the district court’s findings, based upon its considered review of the state court record, are relevant to an analysis of both issue and claim preclusion. But because the district court did not address the case through the legal lens of preclusion, and because the underlying state court records were not made part of the record on appeal, we believe the better course is for the district court to take up the preclusion question in the first instance, on remand.
IV
We appreciate that when plaintiffs file lawsuits in federal court after losing in state court, district courts may naturally and appropriately anticipate that the federal case could fail procedurally, based on various doctrines designed to promote finality of litigation and respect for state courts. Many cases in which Rooker-Feldman is raised involve determined litigants, sometimes without legal representation, who may offer prolix allegations concerning not just the underlying facts but the lengthy history of their prior litigation efforts. We understand that working through these cases can present challenges for district courts. And in this case, although we conclude that the district court erred in dismissing for lack of jurisdiction under Rooker-Feldman, we commend the district court for conscientiously examining the state court record and the various iterations of the plaintiffs’ complaint, in the face of allegations that were sometimes unclear.
The teaching of Exxon Mobil, Noel, and the other precedents in this area is that district courts encountering lawsuits brought by disappointed state court litigants may not regard the fact of a state court judgment as reason sufficient to dismiss the case for lack of subject matter jurisdiction under Rooker-Feldman. Because federal courts cannot disavow the jurisdiction granted to them, there is an important distinction between dismissal under Rooker-Feldman and rejecting a claim based on preclusion or some other ground. Our decision today reaffirms that delineation.
REVERSED AND REMANDED.
MIROTH V. COUNTY OF TRINITY
VANDYKE, Circuit Judge, dissenting:
VANDYKE, Circuit Judge, dissenting:
I agree with the majority’s statement of the law on the Rooker-Feldman doctrine. But I disagree with its application of that law to the unique circumstances in this case. What makes this case unique is twofold. First, it is clear from reviewing their operative complaint that the gravamen of plaintiffs’ claimed injuries stems from the state court judgment removing their children. Second, and critically, plaintiffs have expressly disclaimed any extrinsic fraud by defendants in the state court proceedings, meaning that no act or omission independent
A.
The majority correctly explains that Rooker-Feldman applies only in limited circumstances, yet the doctrine “has sometimes been construed to extend far beyond [its] contours.” Exxon Mobil Corp. v. Saudi Basic Indus. Corp., 544 U.S. 280, 283 (2005). Properly understood, “Rooker-Feldman prohibits a federal district court from exercising subject matter jurisdiction over a suit that is a de facto appeal from a state court judgment.” Kougasian v. TMSL, Inc., 359 F.3d 1136, 1139 (9th Cir. 2004). A de facto appeal exists when a plaintiff both “asserts as a legal wrong an allegedly erroneous decision by a state court” and “seeks relief from a state court judgment based on that decision.” Noel v. Hall, 341 F.3d 1148, 1164 (9th Cir. 2003). But we have never construed Rooker-Feldman so narrowly that it applies only in those situations where the plaintiff expressly asks a federal court to overturn the state court decision. To the contrary, we have acknowledged that the doctrine applies to “forbidden” challenges that are “de facto appeal[s] … ‘tantamount to … moving to set aside a judgment.’” Kougasian, 359 F.3d at 1139–41 (quoting Barrow v. Hunton, 99 U.S. (9 Otto) 80, 82–83 (1878)). The use of words like “de facto” and “tantamount” in reference to the doctrine would be unnecessary if Rooker-Feldman applied only to express requests to reverse a state court. See Noel, 341 F.3d at 1155 (“Rooker-Feldman becomes difficult—and, in practical reality, only comes into play as a contested issue—when a disappointed party seeks to take not a formal direct appeal, but rather its de facto equivalent, to a federal district court.”). Instead, cases have focused on whether the two conditions recited in Noel are met, without imposing a specific words requirement. See, e.g., Kougasian, 359 F.3d at 1139–40 (finding one of the two conditions unmet after thoroughly evaluating the actual source of the alleged legal wrong). Or put slightly differently, in applying Rooker-Feldman, we have asked whether the federal suit is effectively a request for the federal court to remedy harm caused by the state court’s decision, not just whether the federal plaintiff is expressly asking us to reverse the state court.
Here, while plaintiffs have amended their most recent complaint to remove their prior requests for relief expressly directed at the state court, their federal claims continue to demand relief that is predicated on the state court having committed “a legal wrong” against them by ordering the removal of their children, and “seek[ing] relief” from the results of that “state court … decision.” Noel, 341 F.3d at 1164. And plaintiffs have made clear they are not claiming that some extrinsic fraud by defendants prevented them from presenting their case to the state court. Their federal claims are therefore barred by Rooker-Feldman.
B.
The majority provides a few reasons for its different conclusion. To begin, it emphasizes that plaintiffs are no longer expressly asking the district court to reverse the state court judgment. True, plaintiffs now only request monetary damages to
Notably, plaintiffs initially asked the district court to set aside the state court judgment by demanding that the federal court reinstate their parental rights. This obviously violated Rooker-Feldman. After the district court properly dismissed their earlier complaint for that reason, plaintiffs amended their complaint to drop their direct requests to reverse the state court. But the thrust of plaintiffs’ allegations remains unchanged: they continue to seek a remedy for the state court’s termination of their parental rights. If dropping the explicit request to set aside a state court judgment controlled whether Rooker-Feldman applies, a de facto appeal could proceed whenever a party functionally—but not expressly—challenges the state judgment. Rooker-Feldman would essentially be reduced to a matter of careful pleading. To my knowledge, no court has ever characterized the doctrine as simply a pleading requirement. Quite the opposite: “Can a federal plaintiff avoid Rooker-Feldman simply by clever pleading … by alleging that actions taken pursuant to a court order violate his rights without ever challenging the court order itself? Surely not.” Hoblock v. Albany Cnty. Bd. of Elections, 422 F.3d 77, 88 (2d Cir. 2005); see also Klimowicz v. Deutsche Bank Nat. Tr. Co., 907 F.3d 61, 65 (1st Cir. 2018) (“Nor can the plaintiff evade the reach of the Rooker-Feldman doctrine by artful pleading.”); May v. Morgan Cnty., 878 F.3d 1001, 1005 (11th Cir. 2017) (“Although narrow in its application, a state court loser cannot avoid Rooker-Feldman’s bar by cleverly cloaking her pleadings in the cloth of a different claim. Pretext is not tolerated.”).
For Rooker-Feldman to serve its narrow but important purpose, a party’s creative word choice can’t artificially cabin the doctrine. The district court in this case correctly recognized that, “in effect, plaintiffs are asking the court to scrutinize the decisions of the state courts” in order to obtain “relief from the consequences of those judgments.” At its core, plaintiffs’ amended complaint still “seeks relief from a state court judgment.” Noel, 341 F.3d at 1164. Rooker-Feldman applies because “‘the injury of which [plaintiffs] complain[] was caused by the [state court] judgment, … [and] [plaintiffs] did not suffer an injury out of court and then fail to get relief from [the] state court; [their] injury came from the [state court] judgment.’” Id. at 1164–65 (quoting GASH Assocs. v. Vill. of Rosemont, 995 F.2d 726, 729 (7th Cir. 1993)). The relief they seek falls squarely within “cases of the kind … brought by state-court losers complaining of injuries caused by state-court judgments … and inviting district court review and rejection of those judgments.” Exxon, 544 U.S. at 284. Omitting their express request for reversal did not meaningfully transform either the source of plaintiffs’ injury—the state court judgment—or the fact that they are asking the federal court to remedy that injury. Where, as here, a plaintiff seeks such a functional reversal of the state court, Rooker-Feldman applies.
C.
The majority also finds dispositive plaintiffs’ recharacterization of the asserted legal
Looking at the thrust of plaintiffs’ causes of action in this case, the asserted injuries clearly result directly from the state court’s judgment. In their first three claims, for example, plaintiffs contend that defendants made false statements to the state court on which that court relied. But none of those statements injured plaintiffs apart from the
state court’s judgment.1 Plaintiffs were injured only because the state court relied on those statements in ordering the removal of plaintiffs’ children. After considering all the evidence and arguments from both sides, the state court—not the defendants—ordered that plaintiffs’ children should be removed. It is that decision by the state court that caused the injury that plaintiffs still seek to remedy in federal court.
To be fair, there are a few factual allegations scattered throughout plaintiffs’ amended complaint that could be viewed as direct injuries resulting from a source other than the state court. For example, plaintiffs argued that child abuse investigators failed in “the[ir] duty to investigate information that would clarify matters prior to separating children from their parents.” The same investigators also allegedly failed their “legal duty to provide a safety plan.” Be that as it may, plaintiffs have not asserted any independent claims of injury based on these allegations. Cf. Noel v. Hall, 341 F.3d 1148, 1164 (9th Cir. 2003) (“‘The Rooker-Feldman doctrine asks … [if] the federal plaintiff [is] seeking to set aside a state judgment, or does he present some independent claim.’” (quoting GASH Assocs. v. Vill. of Rosemont, 995 F.2d 726, 728–29 (7th Cir. 1993))). Instead, these alleged failures to investigate and provide a safety plan were made in service of the claims regarding the removal of plaintiffs’ children. And plaintiffs’ requested relief confirms they sought no standalone relief based on defendants’ failed duties. So these allegations do not supply a way around
Rooker-Feldman because these injuries necessarily travel through the state court’s judgment ordering the removal of plaintiffs’ children.
That defendants supposedly misled the state court during court proceedings to reach an allegedly wrong decision doesn’t change the reality that plaintiffs’ relief “is contingent upon a finding that the state court decision was in error.” Cooper v. Ramos, 704 F.3d 772, 782 (9th Cir. 2013). Again, it will always be true, in any case where plaintiffs are “alleging injury caused by a state court judgment,” Noel, 341 F.3d at 1165 (citation omitted), that the same injury could also be recharacterized as resulting (even if less directly) from the plaintiffs’ state court adversaries wrongly convincing the state court to issue
Our court has recognized only one exception to Rooker-Feldman where, as here, the state court loser is harmed by the state court judgment. Reusser, 525 F.3d at 859 (collecting cases). Where that harm results from some extrinsic fraud by the state court winner who prevented the state court loser from even being able to present his case to the state court, Rooker-Feldman does not apply. Examples of such extrinsic fraud include hiding the identity of witnesses from your opponent or the state court, see Kougasian v. TMSL, Inc., 359 F.3d 1136, 1138 (9th Cir. 2004), or failing to make mandatory disclosures to your opponent, see Benavidez v. Cnty. of San Diego, 993 F.3d 1134, 1143–44 (9th Cir. 2021).
Extrinsic fraud is distinguishable from a plaintiff claiming intrinsic fraud—that is, where the state court loser instead claims that the defendant directly misled the state court, perhaps by perjury or misrepresenting facts in court.
See Intrinsic Fraud, Black’s Law Dictionary (12th ed. 2024) (providing examples such as the “use of fabricated evidence, perjured testimony, and false receipts”). Intrinsic fraud is “distinguished” from extrinsic fraud because the former goes “to the very heart of the issues contested in the state court action,” and because that type of fraud happened before the state court where plaintiffs “clearly had an opportunity to present their claim” and rebut the supposed fraud. Green v. Ancora-Citronelle Corp., 577 F.2d 1380, 1384 (9th Cir. 1978). Unlike extrinsic fraud, claims of intrinsic fraud are not an exception to the Rooker-Feldman doctrine, presumably because in every case where a federal plaintiff is claiming injury directly from a state court judgment they could just as easily recharacterize that injury as resulting from their state court opponent “misleading” the state court into ruling against them. Such an exception to Rooker-Feldman for intrinsic fraud claims would entirely swallow the rule.
Plaintiffs’ claims in this case are all based on intrinsic fraud, not extrinsic fraud. We know that because plaintiffs in the litigation below disclaimed that they were alleging any extrinsic fraud by defendants. The majority nonetheless refuses to apply Rooker-Feldman because plaintiffs’ amended complaint “does not assert ‘as a legal wrong an allegedly erroneous decision by a state court,’ but rather ‘an allegedly illegal act or omission by an adverse party.’” But this ignores the important difference between claims of intrinsic fraud and extrinsic fraud. It is true that every claim of intrinsic fraud can be characterized as a legal wrong committed by an adverse party. But when you do that and then so rely as the majority does here to conclude that Rooker-Feldman is inapplicable, you effectively eviscerate Rooker-Feldman because every injury directly resulting
from a state court judgment also results from the adverse party having (allegedly) misled the state court into issuing that judgment.
Absent some extrinsic fraud that kept the state court from fully hearing plaintiffs’ claims—which again, plaintiffs have expressly disclaimed here—the exception to Rooker-Feldman does not apply. No matter how plaintiffs recharacterize their federal case, it is still the state court’s removal of their children that caused the relevant injury plaintiffs now seek to remedy. The district court thus correctly found “withdrawal of the extrinsic fraud … dispositive here.”
D.
The majority allocates a lot of space to opining about matters with which I don’t
First, the majority characterizes my position as so broad that it “would rewind the clock to the bad old days in which some courts … improperly applied Rooker-Feldman whenever the federal suit would cast doubt on a state court decision.” Not so. My position is very narrow: when a litigant’s asserted injury flows from the state court’s decision and the litigant expressly disclaims that extrinsic fraud influenced that decision, then Rooker-Feldman applies. Indeed, my position is so narrow that I question whether it would ever apply to a future case because it will be the exceedingly unusual circumstance where a litigant would expressly disclaim extrinsic fraud. It is the disclaimer of extrinsic fraud that does the work for me here. I agree with the majority that plaintiffs “did not walk into a Rooker-Feldman sinkhole the moment they excised ‘extrinsic fraud’
from their complaint.” Plaintiffs walked into that sinkhole the moment they expressly disclaimed extrinsic fraud.
The majority’s attempt to explain why Rooker-Feldman doesn’t apply is not particularly responsive to this unique circumstance. Indeed, the majority’s position essentially collapses extrinsic fraud and intrinsic fraud into one combined category of fraud while ignoring our prior precedents that specifically refer to these distinct types of fraud. See, e.g., Kougasian, 359 F.3d at 1138; Benavidez, 993 F.3d at 1143–44. It can’t be correct that the adjectives “extrinsic” and “intrinsic” have no meaning. And my position is wholly consistent with both Kougasian and Benavidez—where Rooker-Feldman did not apply—because those defendants allegedly procured the state court judgment by extrinsic fraud rather than simply “by fraud.” By excising the adjective “extrinsic” to simply say that Rooker-Feldman does not apply despite the fraud allegation, the majority fails to appreciate the distinct analytical consequences that flow from each type of fraud allegation.
Second, when you cut through the majority’s strawman discussion that is irrelevant to the narrow circumstance on which I’m focused—where a litigant, as here, disclaims extrinsic fraud—the majority’s alarming, and wholly unnecessary, conclusion is laid bare: Rooker-Feldman applies only when a litigant expressly asks to reverse a state court judgment. This is precisely what plaintiffs did in their original complaint. And once plaintiffs amended their complaint to remove the express requests for reversal, the majority contends doing so placed them “outside of Rooker-Feldman for that reason alone.” The majority’s simple articulation of the doctrine notwithstanding, the rule it creates today causes serious tension with cases in our court as well as other circuits.
Start with our court. I’ve already explained that our cases specifically describe Rooker-Feldman as barring “de facto” appeals of state court judgments and efforts “tantamount” to an appeal of a state court judgment. Kougasian, 359 F.3d at 1139–41. But it is worth reiterating that the use of such descriptors for Rooker-Feldman would be pointless if the doctrine applied only to express requests for reversal of a state court judgment. When the majority reads Rooker-Feldman as applying only when a litigant expressly seeks to set aside a state court judgment, it is important to recognize just how remarkable this contention is. To be sure, our court has stressed the narrowness of the Rooker-Feldman doctrine. But we have never before described it as that narrow. See Cooper, 704 F.3d at 777 (“The doctrine bars a district court from exercising jurisdiction not only over an action explicitly styled as a direct appeal, but also over the ‘de facto equivalent’
The majority’s overly narrow articulation that turns Rooker-Feldman into a pleading rule also causes unnecessary tension with other circuits. Other circuit courts recognize that Rooker-Feldman requires an examination of the actual “source of the injury.” Hoblock v. Albany Cnty. Bd. of Elections, 422 F.3d 77, 87 (2d Cir. 2005); see also RLR Investments, LLC v. City of Pigeon Forge, 4 F.4th 380, 388 (6th Cir. 2021); Great W. Mining & Min. Co. v. Fox Rothschild LLP, 615 F.3d 159, 166–68 (3d Cir. 2010). Such an examination would be wholly unnecessary if the rule was truly as the majority posits: that a litigant must expressly request the reversal of a state court judgment. That would always be an easy inquiry and would not have caused the doctrine to “develop[] a reputation for complexity.” See RLR Investments, 4 F.4th at 387–88 (“Usually Rooker-Feldman cases are complicated because it’s difficult to determine if a plaintiff seeks review of a state-court decision … or if a decision counts as a judgment…. ‘But there’s no complexity when the litigant directly asks a federal district court to declare a state-court order to be unconstitutional and enjoin its enforcement.’” (citations omitted)).
The Second Circuit has further explained how this “source of the injury” inquiry would operate in a circumstance similar to the one before us:
Suppose a state court, based purely on state law, terminates a father’s parental rights and
22. The same is true here: plaintiffs contend the state court relied on misleading and fraudulent statements to reach the harmful decision to remove their children. Because, just like in Cooper, “there is already a forbidden de facto appeal” in plaintiffs’ first three causes of action here, “the ‘inextricably intertwined’ test come[s] into play” for plaintiffs’ fourth and eighth causes of action. 704 F.3d at 778 (quoting Noel, 341 F.3d at 1158).
orders the state to take custody of his son. If the father sues in federal court for the return of his son on grounds that the state judgment violates his federal substantive due-process rights as a parent, he is complaining of an injury caused by the state judgment and seeking its reversal. This he may not do, regardless of whether he raised any constitutional claims in state court…. [And] if the state has taken custody of a child pursuant to a state judgment, the parent cannot escape Rooker-Feldman simply by alleging in federal court that he was injured by the state employees who took his child rather than by the judgment authorizing them to take the child. The example shows that in some circumstances, federal suits that purport to complain of injury by individuals in reality complain of injury by state-court judgments. The challenge is to identify such suits.
Hoblock, 422 F.3d at 87–88 (emphasis added). As this example shows, a federal plaintiff can’t avoid Rooker-Feldman by “clever pleading” because federal courts must
judgment authorizing them to take the child[ren].” Id. at 87–88.3
Under the majority’s very narrow view, Rooker-Feldman doesn’t apply here because plaintiffs “are not requesting the return of their children.” But as I’ve explained, the doctrine is not so narrow. It also covers requests for relief predicated on the results of a state court decision. See Noel, 341 F.3d at 1164; Cooper, 704 F.3d at 782. At the end of the day, plaintiffs are seeking money damages as a remedy for the harm produced entirely by the court-ordered removal of their children. The source of that harm is the state court judgment. Even the majority seems to recognize this on some level by acknowledging that “the true ‘consequence[]’ of the state court judgment is that [plaintiffs] lost custody of their children.” Rooker-Feldman does not just preclude express requests that “a federal court … address th[at] consequence[]” by “reinstat[ing] [plaintiffs’] parental rights or vacat[ing] or otherwise revers[ing] the state court orders,” as the majority holds today. For the doctrine to have its intended effect, it must also necessarily preclude requests for money damages designed to remedy the consequences of the state court ordering the removal of their children or terminating their parental rights, which is exactly what plaintiffs complained of here.
The majority thus misses the mark analytically when it says that plaintiffs “are not challenging the actions of the County officials that were ‘pursuant to’ or ‘produced by’ a state court judgment.” Although certain actions by defendants might not have been “produced by” the state court’s judgment, they certainly travel through it and can’t be described as independent claims to avoid Rooker-Feldman’s reach. There is no denying that the fundamental injury in this case—the removal of plaintiffs’ children—was in fact produced by the state court judgment. No amount of reframing the injury as the result of other causes—such as defendants’ actions or arguments preceding the judgment—changes this reality because none of those causes could have led to the removal of plaintiffs’ children absent the state court’s judgment.
Such an overly narrow articulation of the Rooker-Feldman doctrine that ignores the true source of the asserted injury exists on an island. I’m not aware of any circuit that has gone so far to say that the only way for Rooker-Feldman to apply is when a federal plaintiff expressly asks to overturn the state court judgment. It is true that the other circuits, like ours, have appropriately recognized that Rooker-Feldman is a narrow doctrine. But respecting the narrowness of the doctrine doesn’t mean effectively narrowing it out of existence as the majority does here.
Third, the majority contends that it doesn’t reduce Rooker-Feldman to a mere
because “[t]here is nothing new about plaintiffs drafting complaints to account for jurisdictional … impediments.” And even if pleading around Rooker-Feldman was problematic, the majority further writes that off because other procedural mechanisms, like issue preclusion, will fill in the gap. Even if true in practice (which I certainly hope), the availability of another procedural device doesn’t give us license to effectively paper over Rooker-Feldman as the majority does. Indeed, the distinction between Rooker-Feldman and preclusion is important and must be preserved—in large part because the former is jurisdictional while the latter is not. See Lance v. Dennis, 546 U.S. 459, 466 (2006) (“Rooker-Feldman is not simply preclusion by another name.”); Exxon Mobil Corp. v. Saudi Basic Indus. Corp., 544 U.S. 280, 284 (2005) (observing that Rooker-Feldman implicates federal courts’ subject-matter jurisdiction); see also Great W. Mining & Min. Co., 615 F.3d at 170 (“As Exxon Mobil makes clear, the Rooker-Feldman inquiry is distinct from the question of whether claim preclusion (res judicata) or issue preclusion (collateral estoppel) defeats the federal suit. Importantly, preclusion is not jurisdictional.” (citation omitted)).
So what does the majority’s decision mean going forward? Well, state court losers can now bring federal lawsuits challenging unfavorable state court judgments if they do just one of the following: (1) avoid expressly requesting reversal of the state court’s decision or (2) simply allege that their opponent made false and misleading statements notwithstanding the fact that those statements were challenged in front of the state court. And either pleading strategy goes far beyond the narrow focus of my dissent: the exceedingly rare situation where a plaintiff disclaims extrinsic fraud. The majority’s only response to this floodgates concern is to say that another tool, issue
preclusion, will prevent these lawsuits. I hope so. But I’m still not sure why that gives the majority license to reduce the Rooker-Feldman doctrine to a mere pleading standard in this circuit.
* * *
The majority misapplies Rooker-Feldman to this admittedly unique case. Under the majority’s rationale, a losing party in state court could seek to remedy any injury directly resulting from the state court judgment in federal court so long as they recharacterize that injury as resulting from their adversary’s intrinsic fraud misleading the state court, instead of from the state court judgment itself. What’s worse, the majority in reaching this conclusion reduces Rooker-Feldman to a mere pleading requirement that is easily circumvented.
Rooker-Feldman applies narrowly, yes, but it shouldn’t be a paper tiger. If plaintiffs were claiming some extrinsic fraud that “‘prevent[ed] [them] from presenting [their] claim’” to the state court, I would readily agree with the majority that Rooker-Feldman does not preclude federal review. Kougasian, 359 F.3d at 1140 (citation omitted). But plaintiffs disclaimed any extrinsic fraud. And they had ample opportunity in the state court proceedings to counter all the alleged intrinsic misstatements and omissions of which they now complain. The state court considered whatever responses the plaintiffs provided and simply