SPROUSE, Circuit Judge:

The United States Parole Commission (Commission) appeals an order of the district court requiring the release of Jerome Gill, a federal prisoner who had been reincarcerated following a parole revocation. The district court held that the Commission impermissibly considered the severity of a post-parole state criminal violation in determining Gill’s reparóle date and that this, in effect, enlarged his punishment for his original crime in violation of the ex post facto clause, U.S.Const. Art. I, § 9, cl. 3, because such consideration of crime severity was not authorized by the pre-1976 provisions of the Federal Youth Corrections Act (FYCA). We reverse.

Gill pled guilty to bank robbery in 1973 and was sentenced to fifteen years’ federal imprisonment pursuant to the Federal Youth Corrections Act. 18 U.S.C. §§ 5005-5026. He was paroled in December 1976, and, in that year, while on parole, was convicted of two separate state offenses in Maryland — armed robbery and of being a “rogue and vagabond.” He was sentenced in Maryland to consecutive terms of five and two years but was paroled in 1979. Upon parole by Maryland, Gill was taken into federal custody pursuant to a previously-lodged detainer. A federal parole revocation hearing was held, parole was revoked on December 4, 1979, and a presumptive reparóle date of November 30, 1981, was established. The reparole guidelines in force indicated a customary range of 32 to 40 months to be served before re-release, and Gill, under a strict application of those guidelines, would have been re-released immediately since he was given credit for the 38 months in state custody. In setting the later date of November 30, 1981, however, the Commission stated that “a decision above the guidelines appears warranted because [Gill was] a poorer risk than indicated by [his] salient factor score in that [he was] on parole for bank robbery less than 8 months when [he] committed two new offenses including a robbery offense.” The district court on April 9, 1981, granted a writ of habeas corpus directing the Commission to hold a hearing within 14 days unless Gill was released prior to that time. He was released on April 17, 1981.

Initially, we are confronted with a question of mootness. The presumptive parole date set by the Commission has now passed, but the Commission purported to reserve the right to reincarcerate him if it prevails on this appeal. Since Gill has been free on parole for almost a year, the Commission may well decline to reincarcerate *601 him if he has, during that time, demonstrated a capacity to be rehabilitated. Since, however, the possibility remains that Gill can be reincarcerated, the issues raised by the Commission are not moot.

The district court in granting the writ of habeas corpus indicated that the severity of the original offense could not be considered in determining a reparóle date for a prisoner originally sentenced under the Federal Youth Corrections Act (FYCA) for an offense committed prior to the 1976 amendments to that Act. We agree. Prior to 1976, crime severity was not a separate factor to be considered in fixing parole for' youthful offenders. Dorszynski v. United States, 418 U.S. 424, 94 S.Ct. 3042, 41 L.Ed.2d 855 (1974); Marshall v. Garrison, 659 F.2d 440 (4th Cir. 1981); Shepard v. Taylor, 556 F.2d 648 (2nd Cir. 1977). Congress in amending the FYCA, however, explicitly included crime severity as one of the factors to be considered in thereafter determining parole and reparole dates. Marshall v. Garrison, supra, at 443. Despite that amendment, however, the Commission cannot look to severity when setting a reparole date for a prisoner who has committed a crime while on parole, whether before or after 1976, if he committed the crime for which he was originally imprisoned prior to 1976. This would be as much a violation of the prisoner’s right to be free from ex post facto punishment as consideration of crime severity in determining the original parole term. Cf. Shepard, supra, (prisoner sentenced under FYCA in 1972; severity of 1974 offense may not be considered at post-1976 reparole proceeding).

If the Commission had relied on the severity of the crimes giving rise to Gill’s parole revocation as the basis for setting his reparóle date, then it would have been in error. We do not interpret the Commission’s statement, however, as indicating that the severity of the Maryland crimes was the basis for either the parole revocation or the imposition of the additional incarceration. While its formal statement of the reasoning underlying Gill’s reparole date is not a paradigm of clarity, it is apparent that the Commission was concerned with the very fact of Gill’s continued course of criminal activity, and it was the fact of this conduct which caused it to fix the release date it did. Therefore, in this case, the Commission’s formulation of a reparóle date was not impermissible, as its ultimate determination did not rely on the severity of Gill’s post-parole offenses. By basing its decision on an individualized analysis of Gill’s poor conduct after release, the Commission utilized the very criteria embodied in the original FYCA. As stated by the Supreme Court in Dorszynski, supra, “[a]n important element of the program was that once a person was committed for treatment under the Act’ the execution of sentence was to fit the person, not the crime for which he was convicted.” 418 U.S. at 434, 94 S.Ct. at 3048.

Gill contends that his institutional record was good and indicated that he was capable of rehabilitation. A parolee’s institutional record, of course, can be of significant assistance in divining his ability to readapt to a normal life outside of prison, but that is not to say that other evidence cannot be utilized. Realistically, future conduct can never be determined with complete accuracy and, at best, such predictions are hazardous. A parolee’s conduct after he is released, however, is highly probative of his potential for rehabilitation through parole. Shortly after he was paroled, Gill committed two separate and distinct crimes in rapid order. Regardless of severity, whether the crimes had been misdemeanors or violent felonies, the fact of criminal activity rebutted Gill’s exemplary institutional record and indicated that he was a poor risk for rehabilitation through parole.

Since the Commission acted within its statutory authority in considering the fact of Gill’s criminal conduct after his release, and did not set his reparóle date in contravention of the ex post facto clause, the decision of the district court is reversed.

REVERSED.