The borderland between schizophrenia psychoses and affective disorders has always been ill defined. The findings on the `category' of schizoaffective psychoses conflict on the genetic, biochemical, neurophysiological, psychodynamic, descriptive and treatment response as well as on course and outcome level, internally as well as externally; all usual validity criteria for a psychiatric disease. Six hypotheses are evaluated in a literature review. The main conclusions are that schizoaffective disorders form a heterogeneous group of disorders, the affective schizomanias may be related to bipolar affective disorders. In the other postulated subtypes further clinical research to subdivision is necessary in relation to family history. Non biological etiological factors may frequently lead to the occurrence of two diseases especially in schizodepressions.

Next course and outcome and psychopharmacological treatment response are critically evaluated as validity criteria and emphazised is the necessity to check for the presence of symptoms in affective disorders in genetic, descriptive, course and treatment response studies.