A review of the literature suggested a lack of clarity in the distinction between primary and secondary delinquency prevention and early intervention, particularly if the focus of primary prevention was on environmental or institutional change. An analysis of variables associated with predicting youth gang delinquency among inner-city, minority male adolescents identified the family, the school, and the peer group. Gang involvement and delinquency were significantly and positively related for both black and Hispanic youth gang subcultures. Relationships between variables and delinquency and gang involvement, however, varied considerably between black and Hispanic youth. For example, family structure was related to delinquency for blacks but not for Hispanics. Only the presence of a gang member in the family was invariably related to delinquency and involvement for both ethnic groups. The school was a uniquely important institution in terms of social control and the learning environment and their impact on youth gang involvement. Friends and peer groups were significantly associated with gang involvement and delinquency. The presence of gangs in the community was a uniformly important predictor of gang involvement and delinquency for both blacks and Hispanics. While there was a significant relationship between age and gang involvement for Hispanics, there was no such relationship for blacks. For Hispanic youth, gang involvement suggested a time- limited or adolescent period function. Self-esteem measures indicated conflicting spheres of influence within the social world of adolescents. Positive relationships with the peer group and youth gang involvement stood in opposition to negative self-esteem as manifested in the family and school. The analysis suggests that the ideas of gang involvement, delinquency, and nondelinquency have to be distinguished when addressing the role and status of youth gang members and that different gang subcultures prevail in the inner city based on distinctive social conditions and school and peer group influences. Recommendations are offered that focus on treating gang delinquency as a social problem, the need to consider ethnicity, and the development of procedures for measuring gang involvement. Appendixes contain the forms used in the analysis and supplemental data. References and tables