Postmenopausal oestrogen therapy reduces the risk of osteoporosis and cardiovascular diseases but is associated with an increased risk of endometrial cancer. We have assessed the impact of a regimen of oestrogen with cyclic progestagen on risk of endometrial cancer for postmenopausal women. We did a population-based case-control study of women aged 45-74 years in western Washington State, USA. Cases were identified from a regional cancer registry as having histologically confirmed endometrial cancer during 1985-91. 832 (72%) of 1154 eligible cases completed interviews. Controls were identified by random digit dialling, screened for intact uterus, frequency matched for age and county, and randomly assigned a reference date within 1985-91. Interviews with 1114 (73%) of 1526 eligible controls were done. The women provided information about use of hormone replacement therapy, and reproductive and medical history before diagnosis date (cases) or reference date (controls). Relative to women who had never used hormones (for > 6 months), women who had taken unopposed oestrogen had a four-fold increase (95% CI 3.1-5.1) in risk of endometrial cancer. Women who used a combined therapy of oestrogen with cyclic progestagen (eg. medroxyprogesterone acetate) had a relative risk of 14 (1.0-1.9). Among women with fewer than 10 days of added progestagen per month, the relative risk was 3.1 (1.7-5.7). Whereas that for women with 10-21 days of added progestagen was 1.3 (0.8-2.2). The use of these combined regimens for 5 or more years was associated with risks of 3.7 (1.7-8.2) and 2.5 (1.1-5.5), respectively, relative to non-users of hormones. Postmenopausal women who use combined therapy of oestrogen with cyclic progestagen on a long-term basis have an increased risk of endometrial cancer compared with those who are not on hormone replacement, even when progestagen is added for 10 or more days per month. This increase is much smaller than that associated with unopposed oestrogen, but needs to be confirmed.