Fourteen epidemiologic studies of the relationship of rheumatoid arthritis to schizophrenia have been conducted between 1934 and 1985. Twelve of the studies report a lower-than-expected rate of rheumatoid arthritis in populations of schizophrenics. Methodologic weaknesses in the studies are assessed. Nutritional, hormonal, psychosocial, genetic, and immunologic data and theories are briefly reviewed which might explain the epidemiologic results. There is sufficient evidence for the negative association between the two disorders to justify further research.