The Seattle Social Development Project (SSDP) followed a multiethnic urban sample of 808 children from the time they entered the fifth grade in 1985. The sample was divided into Gang Members, Youths with Delinquent Peers, and Youths with Non-Delinquent Peers. A consistent pattern of offending was found across the three groups for all 11 delinquency and substance abuse measures. Youth with non-delinquent peers had the lowest rates of offending; youths with delinquent peers had higher rates; and gang members had the highest rates. A second analysis used structural equation modeling to examine the impact of gang membership on delinquency, controlling for associations with delinquent peers. Gang membership contributed to delinquency above and beyond associations with delinquent peers and prior delinquent behavior. The Rochester Youth Development Study produced strikingly similar results. National implications of these two studies for theory and practice include the finding that gang membership has an independent contributing role in the etiology of delinquency, and preventing youth from joining gangs holds promise for preventing and reducing crime and substance use. Figures, notes, references