Criminal convictions of offenders in the sample were tracked for an average of 10.5 years, beginning with the first conviction offense until 6.5 years after the 1982 release. The study assessed the impact of DJR confinement on postrelease criminal behavior. Offense histories of the sample illustrated the level of chronic and serious delinquency among youth committed to DJR. During the 10.5-year period, the 926 offenders were convicted of a total of 16,341 offenses. Confinement in DJR had a significant incapacitation effect. During the average of 241 days in confinement, offenders committed an average of 0.6 new offenses per year, compared to the baseline rate of 4.4 offenses per year prior to confinement. The drop in offenses during confinement represented an apparent reduction of 2,303 offenses due to incapacitation in DJR facilities. Confinement in DJR residential programs produced a significant rehabilitation effect. The rate of offending dropped by an average of 75 percent after release from DJR. Offenders were convicted of an average of 1.1 offenses per year during the 6.5 years after release, compared to 4.4 offenses prior to release. DJR confinement reduced but did not eliminate criminal activity among the offenders. Three variables provided the best combined prediction of recidivism: age at release from DJR, number of convictions prior to release, and ethnicity. A profile of the study sample and an 8-state comparison of recidivism rates are included. Offense type definitions are provided in an appendix. 17 footnotes, 17 tables, and 5 figures