The study sought to determine any program factors linked to increased or decreased favorable postrelease outcomes. The study focused on 16 programs for men and 4 for women. Although all the programs used a cognitive-behavioral treatment approach, variation across programs was expected to occur due to such variables as staff experience and program emphases and policies related to rule enforcement. The outcome variables measured were recidivism and postrelease drug use over a period of up to 3 years following release. These data were obtained from telephone interviews with probation officers conducted at 6 months, 18 months, and 3 years, unless revocation or completion of supervision occurred before the 3-year limit. Recidivism was defined as arrest for a new offense or revocation of supervision; arrest for a new offense only for those released to supervision by a probation officer; or arrest for a new offense for all individuals, whether or not they were released to supervision. Discrete time proportional hazards regression models showed that, after controlling for individual characteristics, no differences were found in outcomes for the men; however, participants in one of the four female programs had significantly higher drug-use rates, and females who participated in another program had significantly lower recidivism rates. These findings provide some support for the premise that cognitive-behavioral treatment, which has generally proven effective, can produce positive results across a number of programs, despite some program variation. Differences for the women across programs may be due to greater variation in program implementation. 3 tables, 5 notes, and 55 references