From within a social learning framework, the authors test the relative contributions of socialization, group pressure, social selection, and rationalization processes in explaining the relationships between peer-group associations, attitudes, and serious theft. Estimating a nonrecursive model, the authors argue that serious theft is as likely to affect delinquent associations and attitudes as associations and attitudes are to affect serious theft. The authors use panel data over three time periods from the National Youth Survey to estimate simultaneously the influence of each social process over time. In addition, the authors estimate a measurement model to correct for the potential biasing effects of correlated measurement error. Estimation of the nonrecursive model reveals that the socialization effects of peers are negligible and insignificant. The group pressure effects are of relatively equal importance compared to the social selection effects; however, the group pressure effects are substantially smaller in magnitude compared to those commonly reported in recursive studies. 4 figures, 6 tables, and 60 references