To test the antitumor effect of gonadotrophin-releasing hormone (GnRH) analogs, 32 consecutive patients with recurrent endometrial cancer that had progressed through conventional treatments were entered into an open observational trial of treatment with this class of compounds. Patients recruited had progressive, symptomatic, and measurable disease. Treatment was with monthly subcutaneous injections of GnRH analog. Measurements of gonadotrophins, sex hormones, and tumor dimensions were made together with repeat biopsy when possible to assess response to treatment. An objective response was seen in nine patients (28%, 95% CI 13-43%). Responses were seen within the first 2 months of treatment and included pelvic as well as distant sites of recurrence. Significantly greater response rates were seen in previously irradiated sites when compared with nonirradiated sites of recurrence (0.01 > P > 0.001). There was no significant difference between the response in patients with G3 lesions compared with patients with G1/G2 lesions (P > 0.5). Response did not correlate with previous progestogen exposure. No evidence of disease flare or drug toxicity was observed. GnRH analogs have a significant and durable antitumor effect in recurrent endometrial cancer which warrants further examination in comparison with progestogens.