OBJECTIVE To study progestin administration at less than monthly intervals in postmenopausal women given continuous estrogen replacement. METHODS Eighty postmenopausal women received 0.625 mg/day of conjugated equine estrogens for 48 weeks. Using a double-masked design, the subjects were randomized to medroxyprogesterone acetate 10 mg/day for 14 days every 28 or 84 days, or the same dosage for 28 of 84 days. Bleeding patterns, endometrial histology, and serum lipids were assessed. RESULTS The total days of bleeding during the 48-week study were significantly reduced (P < .05) in the women given the progestin for 14 days every 3 months (mean +/- standard deviation 29 +/- 16 days) than with the other two regimens. In all groups, secretory endometrium was reported in 17-39%. At 24 but not 48 weeks, simple hyperplasia was observed in one subject in each of the less than monthly progestin groups. Significant increases (P < .05) of high-density lipoprotein cholesterol were observed before but not after medroxyprogesterone acetate in the women receiving it less than monthly. No change was seen with monthly progestin. CONCLUSIONS In this direct comparison, medroxyprogesterone acetate given for 14 days every 3 months elicited less vaginal bleeding than standard monthly administration. Only a single woman had simple hyperplasia with each regimen of progestin given every 84 days. Medroxyprogesterone acetate given for 14 days every 3 months represents a possible alternative to standard monthly therapy if coupled with regular assessment of the endometrium.