It is essential to learn as much as possible from justice interventions—even those that do not appear to be successful. Data came from a sample of youths participating in drug courts in nine sites across the United States and a comparison group of probationers (N = 1,372). Measures were drawn from case records. Path models with direct and indirect effects were analyzed. Aspects of the juvenile drug court process appear to heighten the likelihood of youth failure in the program and recidivism. The ratio of incentives to sanctions was protective as drug court youth who experienced more of the former had a reduced likelihood of recidivism. The article concludes that it is important to examine mechanisms that impact the success of justice interventions.