This is done by examining the intervening processes as specified by, and that are unique to, each theory. Using data from a large-scale nationally representative probability sample, the analysis examines differences in the likelihood of offending based on the reasons for dropping out and across the two social class groups (distinguished in terms of their position relative to the poverty line). Results support the position that dropping out is more likely to be associated with higher levels of involvement in delinquency for middle-class than for lower-class youth. Also, support for strain theory or social-control theory is dependent on the reasons for dropping out of school. These findings suggest that different approaches to drop-outs based on the youth's social class are warranted. Enhancing the graduation rates for middle-class youth may lead to a reduction in offending. 6 tables and 28 references