It is suggested that a large sample of people exhibiting the schizophrenic syndrome be “repetitively sifted” in an attempt to find etiologic and pathogenic subgroups. For this purpose, epidemiological, biochemical, neurophysiological, psychodynamic, genetic, and other “sieves” should be used. Instead of expecting one factor to hold true for all schizophrenics, one might then hypothesize that a viral factor is the relevant one for a fraction of those born in inclement months, while a pathogenic life history might play the primary role in another fraction. Even within groups, one might find fractions, such as patients for whom the dopamine hypothesis is supported, and others for whom it does not seem to play a role. It is unlikely that one single factor plays the necessary and sufficient role in anyone with the schizophrenic syndrome. Therefore, the method of repetitive sifting should be combined with a “mini-max” model: That is, not only should the group as a whole be sifted, but each individual patient should be searched for different contributing factors, and then a rank-order system of the importance of various factors for each given patient should be established. Such a method of identifying multiple etiologies and pathogeneses would then be a rationale for prevention, treatment, and prognosis.