The present study used data from a randomized controlled trial on brief interventions with adolescents to identify distinct longitudinal patterns of substance use and identify predictors, as well as outcomes associated with those use patterns. Data were originally collected for the purpose of evaluating two brief intervention conditions with adolescents who had been identified in a school setting as abusing alcohol or other drugs (total sample, N = 315). Adolescents were randomly assigned to a two-session adolescent-only brief intervention (BI-A), a two-session adolescent-plus an additional parent session (BI-AP), or an assessment-only control session (CON). We located 74 participants to assess them at approximately 3.5 years post-intervention. Three distinct cluster patterns were identified, including a low decreasing, moderate increasing, and high decreasing pattern of use. The low decreasing cluster was associated with the BI-A condition, mono-substance use, and comorbid anxiety symptoms at baseline. The moderate increasing cluster was associated with the BI-AP condition, polysubstance use, and comorbid conduct disorder symptoms at baseline. No variables were found to be predictive of membership within the high decreasing cluster. There were also no differences found between clusters on adjustment outcomes in young adulthood. Overall findings from this study support the long-term efficacy of a brief intervention, without parent involvement, for adolescents experiencing mild to moderate substance abuse problems. Findings also highlight the importance of early intervention and the tailoring of interventions to meet the unique needs of adolescents.