Results of the study show that, within the context of the contemporary sample of previously delinquent youth, drugs compared with alcohol have a more powerful and sustained effect on life-course patterns of criminal offending. Results support the assertion that drug use exerts unique effects on desistance processes, once levels of alcohol use are taken into account. Data also revealed that, in the aggregate, adult social bonds did not have a significant long-term effect on criminal involvement and that social bonds did not significantly mediate the relationship between drug and alcohol use and offending. Although alcohol has been shown to have significant effects on criminal offending, it is argued that drug use and the drug culture in which many contemporary offenders are enmeshed have consequences that often complicate desistance processes in ways that alcohol does not. This study examined the impact of drug use on criminal behavior over a longer span of time and by introducing social network measures as potential mediators. The study relied on three waves of structured interview data derived from a sample of youth originally incarcerated in institutions for delinquent youth in 1985 and subsequently reinterviewed as adults in 1995 and 2003. It assessed and compared the influence of drug and alcohol use on the likelihood of exhibiting a pattern of stable desistance, stable persistence, or an unstable/episodic pattern of criminal involvement. Tables, references