The role of reproductive factors in endometrial cancer risk has been analyzed in a case-control study conducted since 1983 in the greater Milan area on 568 women (cases) with histologically confirmed endometrial cancer and 1925 women (controls) who were admitted for acute, nonmalignant, hormonal, gynecologic conditions to hospitals that cover a comparable catchment area. Compared with nulliparous women, parous women had a 30% lower risk of endometrial cancer, but there was no evidence of a decline in risk with increasing number of births. The risk of the disease decreased with number of spontaneous or induced abortions; the multivariate relative risk estimates were, compared respectively with no spontaneous or induced abortions, 0.5 for women with two or more spontaneous abortions and 0.3 for women with two or more induced abortions; both trends in risk were statistically significant. When parous women only were considered, no association emerged between endometrial cancer and age at first birth, but the risk decreased with increasing age at last birth: compared with women whose last birth occurred before age 25, the relative risk was 0.5 for women who were greater than or equal to 35 years old at last birth, and the multivariate trend in risk was statistically significant. For most of the reproductive factors that were considered, the risk estimates tended to be greater at younger age or among premenopausal women and to flatten off in subsequent strata of age. An association between endometrial cancer and age at first birth was observed in women who were less than or equal to 49 years old, but not in older groups. The observation that later age at last birth as well as later first birth in younger women decreases the risk of endometrial cancer suggests a short-term protective effect of pregnancy. This finding is consistent with a late-stage (promotional) effect of reproductive factors on endometrial carcinogenesis.