Fifty psychiatric inpatients with a prolonged length of stay were compared to 50 control admissions for factors associated with prolonged hospitalizations in a general hospital. Seven variables were found to be significantly overrepresented among the long stayers, including treatment with electroconvulsive therapy, medical consultations, underemployment, dementia, disposition to a place other than home, absence of alcohol or drug abuse, and presence of psychosis without affective symptoms. The clinical and policy implications of these finding are discussed.