Core research questions focused on differential effects of specific interventions on delinquent behavior and the change process involved in such behavior. In comparing alternative treatments for troubled youth, results indicated that different treatment modalities were implemented with integrity and comparable strength. The three contents of intervention (action, action family focus, and relationship) were equally effective in reducing recidivism. In general, these contents were superior to nonspecific attention intervention (attention-placebo) and treatment-as-usual controls. Activities related to the differential reinforcement of prosocial behavior and social involvement were superior to treatment-as-usual and attention-placebo controls and suggest that interpersonal contingencies, social opportunity, and social involvement are functionally related to delinquency. There was also some support for social labeling notions and for the belief that mere diversion from the court to alternative services may reduce recidivism somewhat. Different change agents were equally effective in reducing recidivism, indicating that social distance may not be a potent contributor to delinquent outcomes. Efforts to select change agents to work with troubled youth were not all successful. Change agents were obviously involved in a tightly controlled and monitored intervention and were highly self-selected and relatively homogeneous. When self-reported delinquency was used as the outcome criterion, instead of recidivism, none of the research interventions showed differential efficacy. Theoretical models of juvenile delinquency and intervention and an empirical history of treatment efficacy are presented. Appendixes contain the forms used in the two studies. References, tables, and figures