OBJECTIVE: The authors sought to identify risk factors for rehospitalization in a seriously mentally ill population, focusing on factors that have the potential to be modified through community-based interventions. METHOD: A case-control design was used in which 101 "case" subjects (recently readmitted psychiatric patients) and a comparison group of 101 subjects living in the community who had been previously hospitalized at the same time as the case subjects, but who in contrast had not been readmitted, were matched on gender, ethnicity, and length of time at risk for rehospitalization. The setting was the Mississippi public mental health system during the first 3 months of 1988, including Mississippi State Hospital and the 10 community mental health regions in its catchment area. The subjects were between the ages of 18 and 55 years, had had at least one previous Mississippi State Hospital admission, and had a primary chart diagnosis of schizophrenia; 197 informants, mostly family members, were also included in the study. Data were collected from structured interviews of subjects and informants, direct observation ratings of subjects, Mississippi State Hospital administrative records, and community mental health center administrative records. RESULTS: Medication noncompliance, comorbid alcohol abuse, and a high level of criticism of subjects by informants were associated with greater risk of rehospitalization, while types and extent of outpatient service use, access to care, quality of life, and demographic variables (other than ethnicity and gender) were not. CONCLUSIONS: These findings imply that interventions aimed at improving medication compliance, reducing alcohol abuse, and helping families cope with their mentally ill relatives could reduce the risk of hospitalization in this population.