Despite the proliferation of drug courts and other mandatory treatment models, few studies have compared the impact of different program features comprising these models. This study compared three groups of clients (N = 330) mandated to the same long-term residential treatment facilities. Study participants were referred from two highly structured programs or from more conventional legal sources, such as probation or parole agents. Analyses showed that these clients varied substantially in their perceptions of legal pressure, and these perceptions generally corresponded to the programs' different coercive policies and practices. Retention analyses confirmed that the odds of staying in treatment for six months or more was nearly three times greater for clients in the most coercive program compared to clients in the third group. Results support the use of structured protocols for informing clients about legal contingencies of participation and how that participation will be monitored, and developing the capacity to enforce threatened consequences for failure.