Patients (173) of mixed diagnoses were randomly assigned to short-term (ST: mean 89 days) or long-term (LT: mean 179 days) hospitalization. Treatment consisted of intensive milieu therapy, individual psychotherapy 2 to 3 times/wk, medication, group therapy and family therapy when deemed appropriate by the hospital staff. LT patients received group therapy less often. Of the patients, 126 were followed-up 3 to 4 yr after discharge. The 2 groups did equally well, but of the patients who were rehospitalized during the first 3 postdischarge yr (30%), LT patients were rehospitalized significantly more often and for longer periods. ST patients received more group therapy after discharge while LT patients had more private therapy after discharge; the total amounts of outpatient therapy did not differ, which indicates a tendency for patients to continue the type of treatment they received in the hospital. ST patients more often received antidepressants after discharge and felt they received more benefit from the hospitalization. A beneficial effect of inpatient group therapy in terms of postdischarge role functioning was found. LT hospitalization should be avoided whenever possible since it does not lead to greater benefit in any of the evaluated areas of adjustment and appears to predispose to more and longer rehospitalization. The possibility that different postdischarge treatments affected the LT-ST comparison on the other outcome variables was not supported.