OBJECTIVE: To determine if estrogen replacement therapy, in women with a history of endometrial cancer, increases the risk of recurrence or death from that disease. METHODS: Two hundred forty-nine women with surgical stage I, II, and III endometrial cancer were treated between 1984 and 1998; 130 received estrogen replacement after their primary cancer treatments and 49% received progesterone in addition to estrogen. Among this cohort, 75 matched treatment-control pairs were identified. The two groups were matched by using decade of age at diagnosis and stage of disease. Both groups were comparable in terms of parity, grade of tumor, depth of invasion, histology, surgical treatment, lymph node status, postoperative radiation, and concurrent diseases. The outcome events included the number of recurrences and deaths from disease. RESULTS: The hormone users were followed for a mean interval of 83 months (95% confidence interval [CI] 71.0, 91.4) and the nonhormone users were followed for a comparable mean interval of 69 months (CI 59.1, 78.7). There were two recurrences (1%) among the 75 estrogen users compared with 11 (14%) recurrences in the 75 nonhormone users. Hormone users had a statistically significant longer disease-free interval than nonestrogen users (P =.006). CONCLUSION: Estrogen replacement therapy with or without progestins does not appear to increase the rate of recurrence and death among endometrial cancer survivors.