It is unclear whether smoking is a risk factor for epithelial ovarian cancer, although some studies have suggested that it may be associated with an increased risk of mucinous tumors. This study investigated the effect of smoking and environmental tobacco smoke (ETS) on ovarian cancer risk among 434 women with primary epithelial ovarian, peritoneal, or fallopian cancers and 868 age- and region-matched hospital controls with nonneoplastic conditions. All participants completed a comprehensive epidemiologic questionnaire. Results indicate that decreased risk of ovarian cancer was associated with being a nonsmoker exposed to ETS (adjusted odds ratio [aOR] 0.68, 95% confidence interval [CI] 0.46-0.99), a former smoker (aOR 0.76, 95% CI 0.53-1.10), or a current smoker (aOR 0.53, 95% CI 0.32-0.88). A similar protective effect was noted for smokers with moderate or high exposure based on smoking intensity, duration, and cumulative exposure, as well as for never smokers exposed to ETS. Results did not differ substantially by histologic subtype. Although prevailing theories of ovarian cancer etiology implicate incessant ovulation, characteristics of the study population suggest that anovulation was not the protective mechanism in this study. Immunosuppression by nicotine or upregulation of enzymes that metabolize carcinogens may be responsible for the effects observed.