We have reported on a demonstmation project aimed at the integration of state and community-based services. Although the integration is still not complete, it appears to have had a promising impact on systemic indicators of continuity of care. The project was undertaken in a naturalistic context with only minimal administrative adjustments and has required no additional funding for clinical pur- poses. It therefore appears generalizable to other settings. Future goals of the project will be focused on further identifying and effectiveby treating patients who do not benefit from community-based services; tracking the course of pa- tients' illness and treatment to devebop capitation payment models; and, finally, working toward more functional administrative relationships with adjacent community and subsuming state systems.