Integrating theories of macro-level social disorganization, micro-level social control, and the life-course/developmental perspective on crime, this study examined the impact of demographic measures (race and gender), neighborhood context, family structure and support, and peer and school associations on offending trajectories. Data for the study were drawn from the 1942 and 1949 Racine Birth Cohort Studies. Longitudinal offending behavior, measured as contacts with the police, were combined with interview data for a subsample of males and females from the birth cohorts. Findings show that five latent classes of offenders are optimal for the 1942 and 1949 Racine Birth Cohort sample as a whole. Five latent classes are also appropriate for the males in the sample, and three latent classes are appropriate for the females in the sample. Both males and females exhibited nonoffending patterns as well as chronic and peaked offending patterns. The low-chronic offenders were the most prevalent offender group and were responsible for the highest percentage of offenses in the cohort (38.8 percent). Although race and gender were important predictors of latent class membership in the baseline models, the addition of social context (neighborhood, peers, and family) rendered them insignificant. Family stability was found to be the most important predictor of chronic offending. Extensive tables and figures and 80 references