OBJECTIVE: This study investigated whether the distinction between schizoaffective disorder and affective disorders with mood-incongruent psychotic features as described in DSM-III-R is reflected by aggregation of schizophrenia in the families of probands with the former disorder and aggregation of affective disorders mainly among the relatives of probands with the latter type of disorders. METHOD: The probands were 118 inpatients with definite lifetime diagnoses of DSM- III-R schizoaffective disorder or a major mood disorder with incongruent psychotic features according to structured clinical interviews. Diagnostic information on 475 of the probands' first-degree relatives was gathered through direct interviews (with 80% of the living first-degree relatives) or the family history approach. The rates of affective and psychotic disorders among these relatives were then compared with those among the relatives of a comparison group of 109 interviewed individuals from the general population who were matched on sociodemographic factors to the inpatient probands. RESULTS: With regard to the familial aggregation of schizophrenia, the DSM-III-R distinction emerged as valid. However, the risk of unipolar affective disorders was enhanced in the families of all of the subgroups of patients studied. The unipolar/bipolar distinction in both DSM-III-R diagnostic groups was reflected by distinct patterns of bipolar disorders in the relatives. CONCLUSIONS: The results partly support the DSM-III-R dichotomy of schizoaffective disorder and affective disorders with mood-incongruent psychotic features. Although the differences between these two diagnostic groups were significant, the magnitude of the differences remained relatively modest.Abstract Teaser