From 237 patients examined for drug-induced psychoses, 50 cannabis psychoses were examined according to the criterion "main cause of addiction" and 107 were examined according to the criterion "consumption during the last three months before hospitalization". The cannabis psychoses were compared to the other drug-induced psychoses as well as to a control group consisting of 219 schizophrenic patients. General agreement was found with the other drug-induced psychoses as well as with the group of schizophrenic patients. The variation from the symptomatology of the schizophrenics is generally common to both the cannabis psychoses and the other drug-induced psychoses. Judging by the results of our investigations, it must be concluded that there is no disease "cannabis psychosis" in its own right, just as the disease "drug-induced psychosis" also does not exist in its own right. While there is a certain slight drug-specific psychopathological undertone, it does not entitle us to speak of a syndromatic or indeed a nosological entity. The psychopathological cross section does not permit a differentiation in the individual psychoses groups mentioned, although this has often been attempted in the literature. That there are no relevant psychopathological differences between cannabis psychoses and endogenous schizophrenia could, for one, be based on the fact that we are observing the final stage of one and the same underlying pathological process. In this case both syndromes would in practice be endogenous psychoses, with the cause not being known in one case. The psychopathologic similarity of these two psychoses forms could, however, also be based on the assumption that cannabis psychoses are triggered schizophrenias, so that we could in both cases be dealing with one and the same disease. We see the solution to the problem of diagnosing symptomatic psychoses, and in particular cannabis psychoses, in making a diagnosis that takes the etiology into consideration in addition to the syndrome diagnosis.