Problems that confound the clinical characterization of polycystic ovary syndrome (PCOS) also complicate the search for its genetic cause. There is no consensus as to the nature of the clinical phenotype for PCOS, and there is even doubt whether polycystic ovaries are an indispensable part of the syndrome. Few ethnic studies on PCOS have been performed, although it has been reported in most major racial groups. Genetic studies of family clusters and relatives of affective probands have shown a high incidence of affected relatives. A dominant mode of inheritance, rather than a recessive one, seems more likely. Multiple genetic causes of adult-onset hyperandrogenism and chronic anovulation have been identified. Chromosomal studies of patients with PCOS have shown no consistent abnormality. Molecular genetic studies are now ongoing. Future genetic models should consider such problematic areas as the heterogeneity of the syndrome, phenotypes in males as well as in females in the nonreproductive years, the exclusion of secondary causes of hyperandrogenism, and the persistence of a syndrome that adversely affects fecundity.