Biological studies of the relationships between the schizoaffective disorders, the affective disorders, and schizophrenia suggest that no simple reductionist model is supported by currently available data. Thus, both affective and schizoaffective patients but not schizophrenics, manifest abnormalities such as decreased platelet serotonin (5-HT) uptake, blunted clonidine-induced increase in serum growth hormone, shortened latency of rapid eye movement (REM) sleep, and increased REM density. However, there are some types of studies which show greater similarity between schizoaffective and schizophrenic patients than between schizoaffectives and affectives--e.g., increased cerebrospinal fluid (CSF) norepinephrine levels, increased platelet 5-HT content, and decreased prostaglandin E1-stimulated adenylate cyclase activity. Other types of studies show abnormalities common to all three groups of psychoses--e.g., eye tracking dysfunction, elevated CSF concentration of gamma-aminobutyric acid, and neuromuscular abnormalities. There are also abnormalities that have been reported to be present in only one type of the psychoses. Although none of these findings have been so unequivocally demonstrated that they can be considered to be firmly established, they do suggest that it is premature to conclude that the schizoaffective disorders are subtypes of the affective disorders. The possibility of a continuum model of the psychiatric psychoses of unknown etiology merits further consideration. Further biological studies of a broad range of psychiatric psychoses with inclusion of the schizoaffective categories appear indicated.