In a prospectively designed study of the course of illness of 161 hospitalized psychiatric patients, data regarding outcome could be obtained for 93% 1 year after clinic discharge. It was possible to reexamine 67% of the patients by means of direct interviews. No significant differences appeared in the comparison of the course of illness outcomes (symptoms, rehospitalization, occupation and social contacts) of patients with schizophrenic psychoses, affective psychoses, neuroses or a group of mixed other diagnoses (predominantly alcohol dependency). For the group of neuroses there was a particularly striking discrepancy between the self- and the observer-ratings at the time of discharge from inpatient index-treatment. From this finding and from the comparatively more intense prominence of depressive symptoms at the time of follow-up, one can presume that there has been insufficient after-care treatment of this patient group considering the recorded treatment data. This seems to hold true for the group of alcohol dependents as well. In contrast, the after-care treatment of patients with affective and schizophrenic psychoses seems more likely to be ensured today. Despite this, however, for the latter the close link between the rate of relapse and the rate of rehospitalization can apparently scarcely be influenced.