The authors describe two psychosocially oriented community residential facilities for patients with persistent and severe mental disorders and multiple failures at community tenure, and they report a retrospective study designed to evaluate treatment outcomes of program residents. The study employed a retrospective single-group repeated-measures design to evaluate 104 patients who completed the one-year follow-up. One-year mean number of admissions to and days in a hospital or crisis center during the two years before program entry were compared with mean admissions and days for the follow-up year; employment status, living status, and Global Assessment of Functioning (GAF) Scale scores at program entry and at one-year follow-up were also compared. Thirteen sociodemographic and clinical variables were individually tested for association with outcome. Hospital and crisis center admissions and days were significantly reduced during the follow-up year. At one-year follow-up, a significantly greater proportion of patients were employed and living independently, and fewer were homeless. GAF scores were significantly higher. No significant correlations between outcome and sociodemographic and clinical variables were found. Despite design limitations of the study, the findings suggest that psychosocial residential treatment models can offer cost-effective and clinically efficacious care to persistently mentally ill patients.