This study examines the rationale for and relative effectiveness of three intervention models for treating people with severe mental illness and substance abuse disorders: Twelve Step recovery, behavioral skills training, and intensive case management. Using clinical trial methods, 132 dually diagnosed clients were assigned to three service approaches. Changes in client psychosocial outcomes, and psychiatric and substance abuse symptomatology were tracked over a 24-month period. Differential effectiveness was evident, with clients in the behavioral skills group demonstrating the most positive and significant differences in psychosocial functioning and symptomatology, compared with the Twelve Step recovery approach. However, the case management intervention also resulted in several positive and important differences compared with the Twelve Step recovery approach. We also found significant changes over time, not only at 6 months but increasingly positive changes in psychosocial functioning at 12 and 18 months as well. These results underscore the need for clinical trials to further examine the relative cost effectiveness of treatment approaches for dually disordered clients and to incorporate means of assessing subgroup differences so that the interventions being tested can be further refined and targeted to a broad set of needs among the dually diagnosed.