We have prospectively studied, by interview, clinical examination and biochemical evaluation, the relatives of 195 consecutive PCOS patients in order to: 1) accurately determine the prevalence of PCOS, as defined by current endocrinological criteria, among first-degree relatives of affected patients, and 2) determine the overall accuracy of proband-only and family member (self report) interview for the detection of clinically evident PCOS within families. We noted that 35% of mothers and 40% of sisters of patients with PCOS will be affected by PCOS themselves. Overall, the interview using a standardized form, whether of the proband or the family relative directly, appears to be a reliable predictor of affected status in mothers. Alternatively, approximately 50% of sisters will be missed using the proband interview, although self-reporting appears to be a reasonably reliable predictor of affected status for these relatives. While we are unable to exclude an autosomal or X-linked dominant mode of inheritance, the heritability of PCOS is probably more complex, similar to that of diabetes mellitus type 2 and cardiovascular disease. In absence of molecular diagnostic markers, a positive family history appears to be the most informative risk factor for the development PCOS.