Projects were implemented in Memphis, Tenn., Newark, N.J.; Boston, Mass.; and Detroit, Mich. Participants were initially placed in small, secure facilities and reintegrated into the community through transitional facilities and then given intensive supervision upon returning to their neighborhoods. Control youths were placed in standard juvenile corrections programs. The underlying principles of program components were the strengthening of personal bonds (attitudes, commitment, and beliefs) through positive experiences with family, schools, the workplace, and nondelinquent peers; the strengthening of social bonds (attachments and involvement) through achievement and successful participation in school, workplace, and family activities; social learning; and goal-oriented behaviors. Program components were case management, the reintegration of youth into their communities, and a multiple-phase residential program that included secure care, community-based residence, and community living. Recidivism and other postrelease data were used for comparing the recidivism and social outcomes of experimental youth with controls. Multiple measures were used for recidivism. The VJOP had little discernable effects on the social indicators of school, work, or family, nor did it strengthen the social bonds thought to be part of the etiology of delinquency. Yet there were indications of reduced recidivism for experimental youths. Although the prevalence of rearrests and reincarceration differed little for experimental and control youth, experimental youths had fewer felony arrests and violent crimes as well as a longer time to first rearrest for all crime types. Policy implications are discussed. 7 tables, 134 references.