Sixty postmenopausal women were enrolled in a 2-year randomized unmasked trial to determine the long-term safety of estradiol (E2) administration by a transdermal therapeutic system. Group I subjects received 0.1 mg of transdermal E2 for 24.5 days of each 28-day cycle for 96 weeks. Group II subjects received the same dosage of transdermal E2 plus 10 mg of medroxyprogesterone acetate, given orally from days 13-25 of each cycle. Vaginal bleeding patterns and endometrial histology were characterized. The subjects recorded bleeding patterns daily. Endometrial biopsies were performed during scheduled follow-up visits at 48 and 96 weeks or as needed to evaluate abnormal bleeding. Data were analyzed by intention to treat. Ten and four subjects dropped out of the study from groups I and II, respectively. A total of 575 and 627 treatment cycles were observed in the same respective groups. Vaginal bleeding was observed in 980 cycles: 381 of 575 cycles in group I (66.3%) and 599 of 627 cycles in group II (95.5%). Bleeding onset, duration, and quantity were similar for both groups. The incidence of hyperplasia was 42 and 4% for groups I and II, respectively, over the 96-week study period. All cases of hyperplasia in group I were treated with sequential medroxyprogesterone acetate for 12 weeks, followed by rebiopsy. In ten of 11 cases, the progestin therapy converted the hyperplasia to a normal endometrium. In one case, the endometrium became hyperplastic again at 96 weeks, but reverted to normal with 12 weeks of medroxyprogesterone acetate.(ABSTRACT TRUNCATED AT 250 WORDS)