The 575 juveniles and their 259 accomplices from a medium-sized community were interviewed at two 6-year intervals. Police records were also examined. The study examined the validity of several hypotheses regarding juvenile crime and development of delinquency. Among the major findings were that most juvenile delinquency consists of gangs of juveniles committing crimes as social activities. Committing crimes together satisfies juveniles' need for togetherness and for excitement. In this respect, criminal behavior does not differ from gangs' noncriminal behavior; committing crime is not the gang's main activity or reason for being. The primary gangs or groups in which crimes are committed have a temporary structure, although the most active delinquents are more likely to continue delinquency while the less active are likely to drop out of the gang after a short period. Members of most of these primary gangs, especially members of the most criminally active ones, are recruited from local social youth environment in which criminality represents an important form of social relationship. The most criminally active juveniles form the cohesive links among criminally active gangs in the same and across communities. The older and more criminally active members instruct younger members in the techniques of crimes as well as in the norms and values associated with juvenile delinquent groups. Young boys tend to be involved with gangs at an early age and drop out of the gang at around age 16. Short association with a gang does not appear to adversely affect the young person's future life. Long association does. For the full report, see NCJ-86378. About 35 references are included.