Data were collected from sample of 182 boys; 151 had no prior placements; robbery was the modal placement offense. The data collection instrument was designed to collect information on intake, placement adjustment, and aftercare. The independent variables used in the study included county of residence, aftercare, partial hospitalization program (PHP) counseling, diploma, risk at intake, risk at release, and placement adjustment. Recidivism, the dependent variable, was defined by rearrests. Statistical analyses provided frequency distributions, while cross tabulations of all independent variables with different forms of the dependent variable demonstrated any significant relationships. The findings show that, despite a higher proportion of high risk cases among residents from Philadelphia as opposed to other locations within the State, there was no statistical differences in recidivism between the two groups. The adapted risk assessment and reassessment instruments predicted recidivism at a statistically significant level. While the PHP counseling program is used more often with high risk cases, there is no difference in recidivism between those who received counseling and those who did not. There was no significant relationship between recidivism and overall adjustment to St. Gabriel's program, however, risk classification at intake was a fair predictor of program adjustment. Group home placement following release seemed to have a positive effect in overriding a poor risk prediction. 2 notes and 2 appendixes