The favourable effects of exogenous progestagen on the endometrium are well known, but have not been adequately quantified with respect to endometrial cancer. The benefits of progestagen need to be weighted against its possible untoward effects on the risk of breast cancer and cardiovascular disease. A population-based case-control study of endometrial cancer was undertaken to evaluate the benefits of progestagen use. 158 incident cases were identified between 1985 and 1987 among women aged 40-64 years who were residents of King County, Washington. Detailed interviews were conducted and the responses were compared with those of 182 controls selected by random telephone digit dialling. The risk of endometrial cancer among women who had used unopposed oestrogen for more than 3 years was over five times that of women who had used no hormones (relative risk [RR] 5.7, 95% confidence interval [Cl] 2.5-12.8), whereas those who had also used a progestagen for at least six months of that time had an RR of only 1.6 (95% Cl 0.6-3.9). The RR differed according to days per month that progestagen was used: 2.4 (0.6-9.3) for progestagen use of less than 10 days per month versus 1.1 (0.4-3.6) for use of 10 or more days per month. These results provide additional evidence that the use of progestagen for 10 or more days per cycle can reduce the excess risk of endometrial cancer associated with long-term postmenopausal oestrogen use.