OBJECTIVE: There is little current understanding of how managed care strategies affect hospital inpatient psychiatric care for mentally ill patients. This study examined one prominent form of managed care, utilization management, which reviews requests for psychiatric care and authorizes provision of care deemed appropriate and clinically necessary. METHOD: The authors analyzed data on 2,265 utilization management reviews conducted during 1989-1992 for patients insured by a single large commercial insurance company. Three utilization management procedures were examined: preadmission review, continued-stay review, and case management. The performance indicators analyzed included percent of admission requests granted, number of days requested and approved, and number of treatment extensions granted. RESULTS: Utilization management initially approved inpatient psychiatric treatment for nearly all (98.8%) of the patients but authorized, on average, only one-third of the days requested (6.9 versus 19.0). On average, 23.5 (total) days of care were requested and 16.8 days were approved. Care for patients with alcohol or drug dependence diagnoses was more restricted than was care for other patients. CONCLUSIONS: These data suggest that managed care does restrict inpatient psychiatric care, primarily by managing length of stay. The fact that almost all patients were approved for the same initial length of stay implies adherence to strict treatment protocols that do not distinguish among different clinical or patient factors. There is a need for careful study of the effects of managed care on outcomes and quality of psychiatric care.