Measures of delinquency and group attributes were obtained through structured interview procedures with a racially mixed sample of youths from two different types of Boston communities. The response rate for the 55 youths with State juvenile authority records was 7.3 percent. For youths who reported having a regular peer group (535 of 589), a standard measure of group solidarity was used. Peers' gang-likeness was statistically, significantly correlated with delinquency, but the correlation was weak and accounted for less than 2 percent of the variation in delinquency. There was no major deviation from this general pattern for males or females in either of the communities. The evidence indicates that peers' organization as a stereotypic gang is not a sufficient condition to stimulate serious delinquency among members. For males in both communities, combinations of the variables -peers' delinquency, gang-likeness and solidarity -- failed to explain significantly more variations in individual delinquency than did these and other peer group variables singly. There was a trend in both communities for a very small number of the highest delinquency males to be concentrated in the gang-like, high-delinquency groups. Some youths who belong to gangs as popularly conceived do break the law and are among the most seriously delinquent of youths, but the present research raises questions about the extent of the coincidence of the seriously delinquent individual and the stereotypic gang.