We now have evidence that the prognosis for schizophrenia is much better in nonindustrial than in industrial societies. This paper reports on a 5-year follow-up of schizophrenic patients living in the peasant society of Sri Lanka and shows that social adjustment and clinical state of a sample of first admission schizophrenic patients examined at the end of 5 years are remarkably good. Further, the 5-year outcome for these patients is consistent with WHO's samples followed in Nigeria and India, for example, and consistently different from outcome for schizophrenic patients followed in industrial societies such as Denmark, U.S.A., U.K., and U.S.S.R. Further, we have shown for the Sri Lanka schizophrenics that good outcome cannot be explained by artifacts of sampling or diagnostic methods, by type of treatment, or by the family's willingness to tolerate deviance. Instead, to explain cultural differences in prognosis, we propose a theoretical alternative to the medical model of disease, social labeling theory, that attributes good prognosis to cultural factors such as the traditional system of beliefs, structure of the treatment system, and family norms. In modern industrial societies, expectations and beliefs about mental illness and the operation of the treatment system serve largely to alienate schizophrenic patients from their normal roles and thus to prolong illness. In contrast, beliefs and practices in nonindustrial societies encourage short term illness and quick return to normality. Cultural differences in prognosis, then, may be the result of culturally based self-fulfilling prophecies.