Little empirical data exist addressing potential iatrogenic effects of placing youth in juvenile justice settings. We took advantage of a natural experiment in one state where juvenile offenders are evaluated in either residential settings characterized by high-density contact with delinquent youth or community settings with naturally varying contact with delinquent peers. Higher rates of subsequent recidivism were found among first-time offenders when evaluation occurred in residential (N = 1,255) as opposed to community settings (N = 752). This finding was replicated in a subset (N = 634 per group) matched using propensity scores for five predictors of recidivism. Findings are interpreted in light of a deviancy training process occurring in residential juvenile justice settings. (Contains 3 tables and 1 figure.)