The PYS is a longitudinal study of the development of delinquency, substance use, and mental health problems among inner-city adolescent males (Loeber, Farrington, Stouthamer-Loeber, and Van Kammen, 1998). Violence was measured annually by the Self-Reported Delinquency scale and included the past year frequency of gang fighting, strong-arming, attacking someone with a weapon or intent to seriously hurt or kill, and rape or forced sex. The self-reported frequency (number of times in the past year) of alcohol, marijuana, and hard drug use was assessed annually with the Substance Use Scale. The analysis focused on the relationship of alcohol and marijuana use with violence; hard drug use was controlled for in some of the analyses. The frequency of alcohol consumption was the sum of the number of times participants drank beer, wine, or hard liquor and did not include trying a few sips or drinking with the permission of adults during special occasions/religious services. Based on the findings of previous studies that have examined substance use and violence, several risk factors common to both substance use and violence were selected as covariates. Risk factors included the lifetime frequency of hard drug use and the frequency of self-reported property crime assessed at age 11. Other child factors were low academic achievement and depressed mood and hyperactivity/impulsivity/inattention problems. Family risk factors included poor communication with caretaker and poor supervision. Longitudinal associations between alcohol and marijuana use and violent behavior were examined first by cross-lagged (year-to-year) associations from ages 11 to 20. Next, several waves of data were aggregated into two time periods that covered ages 11 to 14 and 15 to 20. From ages 11 to 20, substance use was more prevalent and frequent than violence. Substance use continued to increase throughout late adolescence; whereas, violence followed a typical age-crime curve, increasing to approximately age 16 and then decreasing. Frequent alcohol and marijuana use were both concurrently associated with violent behavior. Overall, odds ratios showed that the longitudinal relationship with violence was stronger for marijuana use than alcohol use, and the relationship between marijuana use and violence was bidirectional; earlier violence was also related to later marijuana use, but not alcohol use; however, the association between marijuana use and later violence was spurious; it was mediated by common risk factors. Participants who were African-American or hard drug users were more likely to engage in violence and were also more likely to become frequent marijuana users. The findings do not indicate that early frequent marijuana use causes later violence; rather, frequent marijuana use and violence co-occur because they share common risk factors. The difference between alcohol use's and marijuana use's association with violent behavior can be explained by the fact that alcohol use is accepted, normative behavior, albeit illegal for minors, and marijuana is illegal for all age groups. This makes it more likely that marijuana use will cluster in individuals with more deviant tendencies. Future research should explore the perception that alcohol consumption is more normative among White participants, compared to marijuana use being more normative among African-American participants. 3 tables, 1 figure, and 30 references