LINN, Circuit Judge.

Henkel Corporation (“Henkel”) appeals from a final decision of the Board of Patent Appeals and Interferences (“the Board”), Henkel Corp. v. Procter & Gamble Co., Patent Interference No. 105,174 (B.P.A.I. Mar. 28, 2008) (“Board Decision”), which awarded priority of invention to The Procter & Gamble Company (“P & G”). Because the Board’s decision is supported by substantial evidence, we affirm.

I. BACKGROUND 1

The technology at issue in this case relates to two-region dishwasher detergent *1288 tablets having certain properties. Claim 1 of P & G’s U.S. Patent No. 6,399,564 is representative of Count 2, the only count at issue in the interference:

A detergent tablet comprising a compressed portion and a non-compressed portion wherein:

a) said compressed portion comprises a mould and dissolves at a faster rate than said non-compressed portion on a weight by weight basis, measured using a SOTAX dissolution test method;

b) said non-compressed portion is in solid, gel or liquid form;

c) said non-compressed portion is delivered onto said mould of said compressed portion; and

d) said non-compressed portion is partially retained within said mould; and wherein said non-compressed portion is affixed to said compressed portion by forming a coating over the non-compressed layer to secure it to the compressed portion or by hardening.

At issue in this appeal is limitation a), which requires that the compressed portion of the tablet dissolve at a faster rate than the non-compressed portion. In Henkel Corp. v. Procter & Gamble Co., 485 F.3d 1370, 1375 (Fed.Cir.2007) (“Henkel I ”), we rejected the Board’s overly restrictive interpretation of the limitation and concluded that the limitation required “an appreciation by the inventors simply that the dissolution rate of the compressed region is greater than the dissolution rate of the other region.” We also concluded that Henkel had demonstrated an appreciation of this disputed limitation no later than May 1997. Id. at 1376. Accordingly, we vacated the Board’s denial of Henkel’s motion for priority as well as the Board’s denial (as moot) of P & G’s motions for priority, and remanded to the Board for further proceedings. See id.

On remand, the Board entertained P & G’s motions for priority and held, based on the evidence of record, that “P & G has proved by at least a preponderance of the evidence that it made, and at least one inventor appreciated, an embodiment within the scope of Count 2, the sole count in the interference, prior to Henkel’s earliest reduction to practice.” Board Decision at 3. Consequently, the Board awarded priority of invention to P & G.

Henkel timely appealed. We have jurisdiction under 28 U.S.C. § 1295(a)(4)(A).

II. DISCUSSION

A. Standard of Review

Whether an invention has been reduced to practice is a question of law based on underlying facts. Cooper v. Goldfarb, 154 F.3d 1321, 1327 (Fed.Cir.1998). Accordingly, the Board’s ultimate conclusion of reduction to practice is reviewed de novo, while its underlying factual findings are reviewed for substantial evidence. Henkel I, 485 F.3d at 1374. “Substantial evidence is more than a mere scintilla. It means such relevant evidence as a reasonable mind might accept as adequate to support a conclusion.” Consol. Edison Co. v. NLRB, 305 U.S. 197, 229, 59 S.Ct. 206, 83 L.Ed. 126 (1938).

B. Analysis

Following our remand in Henkel I, the Board found that P & G demonstrated an actual reduction to practice in February 1997, Board Decision at 10, predating Henkel’s reduction to practice in May 1997, and thus awarded priority of invention to P & G, id. at 69. In so doing, the Board made a key factual determination— that P & G inventors appreciated by February 1997 the limitation in the count requiring that the compressed region of the tablet dissolve at a greater rate than the non-compressed region. Id. at 61-62. Whether substantial evidence supports this finding of fact is the only issue on appeal. Henkel contends that “the objec *1289 tive evidence before the Board fails to record or establish that any P & G inventor contemporaneously appreciated, or even conceived of, an embodiment meeting the express comparative dissolution rate limitation of Count 2 prior to Henkel’s accorded actual reduction to practice date.” Henkel’s Opening Br. at 19. P & G counters that its inventors did, in fact, appreciate the differential dissolution rate limitation, and that the record evidence supports the Board’s decision.

“In order to establish an actual reduction to practice, the inventor must prove that: (1) he constructed an embodiment or performed a process that met all the limitations of the interference count; and (2) he determined that the invention would work for its intended purpose.” Cooper, 154 F.3d at 1327. The inventor must also “contemporaneously appreciate that the embodiment worked and that it met all the limitations of the interference count.” Id.; see also Henkel I, 485 F.3d at 1374. As we held in Henkel I, to demonstrate a reduction to practice of the invention at issue here the inventors need only appreciate “that the dissolution rate of the compressed region is greater than the dissolution rate of the other region.” Henkel I, 485 F.3d at 1375.

The Board relied primarily upon three pieces of evidence for its conclusion that P & G inventors reduced the invention to practice prior to Henkel: a Record of Invention by Alasdair McGregor dated November 22,1996 (“the McGregor Record of Invention”), J.A. 321-29, McGregor’s subsequent declaration, id. at 386-92, and a January 1997 Monthly Report by Sabine Metzger-Groom (“the Metzger-Groom Report”), id. at 371; see also Board Decision at 53-54 (“[0]ur focus for actual reduction to practice is on Metzger-Groom’s making and testing of the tablet [described in the McGregor Record of Invention] and [McGregor’s] purported appreciation that the invention worked for its intended purpose.”). In particular, the Board found that although the McGregor Record of Invention does not itself demonstrate an appreciation for the contested comparative dissolution rate limitation, McGregor’s subsequent testimony that he appreciated that limitation at the time, as corroborated by the Metzger-Groom Report, was sufficient to demonstrate appreciation for the comparative dissolution rate and thus to demonstrate an actual reduction to practice. 2 Board Decision at 61-62, 68-69; see also id. at 3.

*1290 Because our review is confined to the question of whether the Board’s determination is supported by substantial evidence, we turn to the evidence the Board relied upon to determine whether it is adequate to support the Board’s conclusion. The McGregor Record of Invention describes a “dimple” tablet — i.e., a compressed tablet with a dimple, into which a molten mixture comprising an inert carrier and a detergent additive is poured and allowed to harden. J.A. 322. The parties do not dispute that the McGregor Record of Invention describes a detergent tablet including all limitations of the count except for the limitation relating to comparative dissolution rates. Although that limitation is absent from the McGregor Record of Invention, McGregor subsequently testified that he appreciated, no later than February 14, 1997, that the compressed region of the tablet dissolved more quickly than the non-compressed region. Board Decision at 14; J.A. 387 ¶ 4. The Metzger-Groom Report documents the preparation and testing of the tablet discussed in the McGregor Record of Invention. In the course of discussing the test results, Metz-ger-Groom states that “the performance of dimple tablets is slightly worse than that of regular tablets,” noting that the loss of performance could be a result of “slower release of NB-base from the dimple vs. regular tablets.” J.A. 371.

The Board read this last statement — i.e., “slower release of NB-base from the dimple vs. regular tablets” — as an appreciation that the compressed region (i.e., the “regular” region) of the dimple tablet dissolved at a faster rate than the non-compressed region (i.e., the “dimple” region). Consequently, the Board concluded that it corroborated McGregor’s testimony that he appreciated the comparative dissolution rates. See Board Decision at 61-62 (“We credit McGregor’s testimony regarding appreciation as it is sufficiently corroborated by, and consistent with, the evidence of record. For example, Metzger-Groom’s monthly report identifies the tested tablets as successfully removing carotenoid stains and as having a slower dissolution rate for the noncompressed region (dimple) as opposed to the compressed region.”).

P & G argues that the Board’s interpretation of the Metzger-Groom Report is correct and that the Board did not err in awarding priority to P & G. Henkel counters, however, with its own interpretation of this evidence. According to Henkel, the statement under examination does not reveal an appreciation of differential dissolution rates of the compressed and non-compressed regions of the same tablet, but instead “implies a potentially slower release of the untablettable active ingredient from the dimple inert carrier than the release of the same active ingredient tab-letted in a regular tablet, i.e., a single-region tablet.” Henkel’s Opening Br. at 11. Henkel therefore contends that the Board erred by awarding priority to P & G.

The interpretation of the critical portion of the Metzger-Groom Report is a very close call, and can reasonably go either way. But our inquiry in this case is not how we would interpret this statement in the Metzger-Groom Report were we to do so in the first instance. Rather, our task is to determine whether the Board’s interpretation is supported by substantial evidence. We conclude that it is. In light of the focus of the Metzger-Groom Report on testing the two-region dimple tablet conceived in the McGregor Record of Invention, as well as the numerous references in the report to dissolution rates in general, e.g., J.A. 371 (noting goal of concept was to “improve stability and rate of delivery ” of cleaning agents (emphasis added)), we find *1291 the Board’s interpretation — that this statement demonstrates an appreciation that the dissolution rate of the dimple, or non-compressed region of the tablet, was slower than the “regular,” or compressed region of the tablet — reasonable. Thus, even if we were to assume for the sake of this appeal that Henkel’s interpretation of the Metzger-Groom Report is also reasonable, the substantial evidence standard of review compels affirmance of the Board’s interpretation. See Guise v. Dep’t of Justice, 330 F.3d 1376, 1381 (Fed.Cir.2003) (finding Merit System Protect ion Board’s interpretation of evidence supported by substantial evidence where the interpretation was reasonable); see also NLRB v. Augusta Bakery Corp., 957 F.2d 1467, 1473 (7th Cir.1992) (“Where two inferences are possible, we cannot substitute our own inference for that of the Board, so long as the Board’s is supported by substantial evidence in the record as a whole.”); Midland Transp. Co. v. NLRB, 962 F.2d 1323, 1326 (8th Cir.1992) (“We may not substitute our interpretation of the evidence for the Board’s reasonable inferences.”).

Because Henkel challenges only the Board’s interpretation of the Metzger-Groom Report and not the Board’s conclusion of corroboration based on that interpretation, we affirm the Board’s award of priority of invention to P & G.

III. CONCLUSION

The decision of the Board is AFFIRMED.

COSTS

No costs.

. This is the second appeal in this case. In Henkel Corp. v. Procter & Gamble Co., 485 F.3d 1370 (Fed.Cir.2007) (“Henkel I"), we discussed at length both the background of the technology at issue and the procedural posture leading to our vacatur and remand of the Board’s decision. We omit the bulk of those details in this opinion, discussing only what is relevant for disposition of this appeal.

. We note that the Board also considered the declaration of David J. Smith ("the Smith Declaration”), which documents Smith's reproduction and testing—many years after the relevant time frame for reduction to practice—of a tablet made according to the McGregor Record of Invention. Board Decision at 22-23; see J.A. at 346-361. According to the Smith Declaration, a tablet made according to the McGregor Record of Invention possesses all the limitations of Count 2, including the relative dissolution rate limitation. J.A. at 347 ¶ 9. Smith’s later recognition of this property of the dimple tablet is irrelevant, however, to whether the inventors contemporaneously appreciated the limitation at the time of the alleged reduction to practice. The Board appeared to recognize as much, noting specifically that it "g[a]ve Mr. Smith’s testimony no weight as to the question of whether the inventors’ [sic] conceived of this tablet property or contemporaneously appreciated this property.” Board Decision at 57. Nevertheless, the Board gave contradictory indications in the same discussion, noting that appreciation of the comparative dissolution rate limitation was "confirmed by the testimony of David Smith,” id., and "crediting] Smith’s testimony, for ... it is consistent with the contents of the [McGregor Record of Invention]," id. The Board's inconsistent and inappropriate treatment of the irrelevant Smith Declaration in this highly fact specific case is not helpful. That fact notwithstanding, the Board’s error is harmless given the Board's explicit disavowal of any reliance upon it. It is clear that what was determinative at the Board was its inter *1290 pretation of the Metzger-Groom Report as being corroborative of McGregor’s testimony.