The study sample was composed of all adult male sex offenders discharged from prisons in the 2 countries in 1979 (n=419). The sample was released at a time when there were no centralized treatment programs for sex offenders. These were not developed nationally until 1992. All offenders in the cohort had been convicted of a sex offense as the cause for their incarceration. One-quarter of the sample had committed a main offense that involved an indecent assault on a female under 16 years old, and one-fifth had a main offense that involved buggery. Sixty-two percent had a previous conviction for an offense, and 24.3 percent had at least one previous conviction for a sex offense. Reconviction data were obtained from official records for the years 1979-2000. The study found that 24.6 percent of the sample was reconvicted for a sex offense over the entire 21 years; 21.7 percent were reconvicted for a violent offense; and 61.8 percent were reconvicted for other offenses. The belief that the risk of reoffending for released sex offenders persists for many years is apparently supported by these findings. Of those reconvicted for a sex offense, 35.9 percent received their first reconviction after at least 5 years in the community. For one-fifth of the sample who received a sex-offense reconviction, 10 years elapsed before their first reconviction. Apparently sex offenders have a longer life span of committing sex offenses than general reoffending. These findings have implications for the length of supervisory periods for released sex offenders. 3 tables, 1 figure, and 24 references