Although peer influence typically has been associated with the promotion of alcohol and drug use among adolescents, a minority of writers has argued that appropriate kinds of peer interactions may work favorably to reduce usage. This empirical study considered two kinds of peer affiliations--formal and informal group participation--and their relationships with use of alcohol and drugs. Findings suggest that ad hoc, informal group participation is related to increased substance use, whereas formal group membership is related to reduced substance use. This article reviews these findings in the context of two theoretical explanations.