• Ninety-one consecutively admitted patients with schizophrenia (n=21), schizoaffective depression (n=43), or psychotic depression (n =27) entered a blind family study along with 36 never-ill controls. Though schizophrenia spectrum disorders clustered within families, they were not significantly more prevalent in the families of schizophrenic probands. In contrast, morbid risks for affective disorder clearly separated the families of psychotically depressed probands from the families of both schizophrenics and controls. Family study data for schizoaffective probands indicated links to both affective disorder and schizophrenia and suggested, as well, that a small number of patients with schizoaffective disorder may carry a genetic liability to both conditions.