Adolescent psychopathology is commonly connected to marijuana use. How changes in these adolescent antecedents and in adolescent marijuana use are connected to patterns of marijuana use in the 20s is little understood. Another issue not clearly understood is psychopathology in the 30s as predicted by marijuana use in the 20s. This study sought to examine these two issues and the associations with marijuana disorder diagnoses using a longitudinal data set of 205 men with essentially annual reports. Individual psychopathology and family characteristics from the men's adolescence were used to predict their patterns of marijuana use across their 20s, and aspects of the men's psychopathology in their mid-30s were predicted from these patterns. Three patterns of marijuana use in the 20s were identified using growth mixture modeling and were associated with diagnoses of marijuana disorders at age 26 years. Parental marijuana use predicted chronic use for the men in adulthood. Patterns of marijuana use in the 20s predicted antisocial behavior and deviant peer association at age 36 years (controlling for adolescent levels of the outcomes by residualization). These findings indicate that differential patterns of marijuana use in early adulthood are associated with psychopathology toward midlife.