The risk of endometrial cancer in relation to cigarette consumption was evaluated in a hospital-based case-control study of breast and genital neoplasms conducted in Milan, northern Italy. For the present analysis, 357 women (cases) with histologically confirmed endometrial cancer were compared to a group of 1122 women (controls) admitted for a large spectrum of acute conditions unrelated to smoking or to any of the known or potential risk factors for endometrial cancer. Compared with never-smokers, the multivariate relative risk estimates were for current 0.45 [95% confidence interval (CI) = 0.30-0.70] and 0.86 (95% CI = 0.50-1.46) for ex-smokers. The negative association of endometrial cancer with current smoking was not influenced by menopausal status as well as by other major identified potential confounding factors, i.e. menstrual and reproductive history, body mass index, oral contraceptive or estrogen replacement therapy use and family gynecologic cancer history. However, there was no evidence of a dose-risk effect, since the relative risks were similar in moderate and heavy smokers. The present study confirms that smoking is less frequent in cases hospitalized for endometrial cancer than in a comparison group of patients with non-smoking-related acute conditions. This negative association is perhaps explained in terms of reduced estrogen levels in smokers, though the influence and the importance of some uncontrolled selection bias (due, for instance, to longer hospital stay of smokers even when admission diagnosis was for non-smoking-related conditions) cannot be ruled out.