The research involved 1,725 youths ages 11-17. The youths were initially interviewed between January and March 1977 regarding their delinquent acts in 1976. The second and third waves were conducted in 1978 and 1979. The cumulative loss of participants over the three waves was 6 percent. Data were collected on a 24-item self-report index that asked participants how often during the last year they had engaged in various delinquent acts. Results suggested that labeling causes delinquency indirectly rather than directly, by means of associations with delinquent peer groups. The measure of delinquent peer associations was a strong predictor of self-reported delinquency. Therefore, the results of the model indicate that many previous conceptualizations of the effects of labeling on subsequent deviance may be incomplete. Findings provide a solid basis for moving to integrate labeling, strain, control, and social learning theories. Future research should determine whether support exists for the overall integrated model. Figure, tables, notes, appended list of the self-reported delinquency behaviors, and 38 references (Author abstract modified)