From the Bonn study including 502 patients followed up an average 22.4 years after onset of schizophrenic disease diagnosed according to the criteria of K. Schneider, we selected 113 cases of schizo-affective, schizophreniform, and cycloid psychoses in accordance to the definitions given by Kasanin, Retterstøl, Angst, and Leonhard. These psychoses have a better prognosis than the whole sample: characteristic residues are seen more rarely, complete remissions and noncharacteristic residues more frequently. This group of psychoses differs from the whole sample in the hereditary taint, too: the morbidity risk with affective psychoses and with schizophrenic psychoses in first- and second-degree relatives is higher than in the total sample of the Bonn study. In spite of the better prognosis and other differences described in the paper, we believe that these results do not justify the classification of schizo-affective and related disorders as an independent disease group. Between these different subtypes of schizophrenia only a differential typology and not a differential diagnostic is possible