The fact that RU 486 curtailed estrogen-induced endometrial proliferation in primates and relieved pelvic pain in women with endometriosis is the reason for continuing research on antiprogestins. Thirty-two adult female cynomolgus monkeys demonstrating menstrual regularity had surgery for the induction of endometriosis. After lesion staging, four treatment groups (n = 8), each of 1-yr duration, were made. Group I received combination/sequential therapy with depot GnRH agonist (GnRH-a) for 3 months, followed by weekly RU 486 for 9 months. Group II received weekly RU 486, group III received monthly GnRH-a, and group IV served as a vehicle control. A staging laparotomy was performed every 3 months to assess the area of peritoneal endometriosis (square centimeters) and the thickness of in situ endometrium. Bone density was measured serially by dual x-ray absorptiometry. Serum was collection weekly. Mean (+/- SE) serum estradiol levels were lower after GnRH-a (77.1 +/- 2.6 pmol/L) than after RU 486 (231 +/- 12 pmol/L) treatment and lower than those in untreated cycling controls (231 +/- 13 pmol/L). GnRH-a produced significant atrophy of endometriotic plaques within 3 months of therapy; this lesion reduction was sustained with RU 486. Both GnRH-a and RU 486 alone produced profound thinning of ectopic and eutopic endometrium throughout 1 yr of continuous therapy. Bone density decreased significantly after 6 months of GnRH-a alone (P < 0.05), without significant changes in the other groups. After RU 486 treatment, there were no significant changes in testosterone, androstenedione, sex hormone-binding globulin, or cortisol. Like GnRH-a, long term antiprogestin therapy produced a reduction in the volume of pelvic endometriotic lesions as well as atrophy of in situ endometrium; however, RU 486 allowed maintenance of tonic ovarian estradiol secretion, suggesting that efficacious endometriosis therapy can be sustained long term without the sequelae of hypoestrogenism, specifically bone density loss.