This prescription-based cohort investigation was undertaken in order to study whether climacteric estrogen treatment of women in a Swedish population might be associated with an increased risk of endometrial cancer, and whether added progestogens can afford any protection from developing estrogen-related endometrial neoplasia. Some 23 000 women who had been prescribed estrogens were followed up regarding the outcome of neoplastic lesions of the endometrium. The results are based on an observation period of one to four completed years, corresponding to 89 000 person-years. Among those cohort members exposed to estrogens alone, regardless of the duration, the relative risk of endometrial cancer was 1.3 (95% confidence interval 0.9 to 1.7). The inclusion of premalignant endometrial changes resulted in a significantly increased relative risk of 1.6 (1.2-2.1). The relative risk estimates in association with estrogen-progestogen combinations were 0.6 (0.2-1.4) for endometrial cancer and 0.8 (0.4-1.5) when including premalignant lesions. These data indicated a possible protective effect of progestogens against the development of endometrial neoplasia. It was concluded that estrogens were - within an observation period of 4 years - associated with an increased risk of premalignant endometrial lesions.