Objective: To determine changes in trabecular vertebral bone mass, serum E-2, and serum calcitonin during and after therapy of pelvic endometriosis with depot leuprolide acetate (LA) or danazol. Design: Prospective, randomized, double-blind study. Setting: Academic university hospital and department of obstetrics and gynecology. Patients: Twelve women with symptomatic pelvic endometriosis diagnosed and staged by laparoscopy. Interventions: All patients received blinded treatment with either 3.75 mg IM depot LA given every month and daily placebo tablets (n = 6) or 800 mg oral danazol daily with a monthly placebo injection (n = 6) for 24 weeks. Main Outcome Measures: Quantitated computerized tomography of bone density of thoracic 12 to lumbar 4 vertebral bodies were determined before, at the end of 24 weeks of treatment, and 6 and 12 months after completing treatment. Gain or loss of bone mass was based against pretreatment levels. Serial serum levels of E-2 and calcitonin before, throughout, and after therapy were compared with changes in bone mass. Results: Bone loss with LA was 14.0% +/- 0.5% (mean +/- SEM), recovering to a deficit of 4.2% +/- 3.8% and 3.3%, 6 and 12 months after stopping therapy. Danazol increased bone by 5.4% +/- 2.2%, with a further gain to 8.2% +/-: 3.5% and 7.5%, 6 and 12 months after stopping treatment. Serum E-2 levels usually were <25 pg/mL (conversion factor to SI unit, 3.671) with LA but >47.3 pg/mL with danazol. Calcitonin levels did not change significantly with either treatment. Conclusion: Depot LA produced marked sustained hypoestrogenemia and significant bone loss with incomplete recovery 1 year after stopping treatment. Danazol maintained normoestrogenemia and increased bone mass with the gain maintained even 1 year after stopping therapy.