Progressive bone loss in osteoporosis results from bone resorption in excess of bone formation. We conducted a double-blind study in 66 women with postmenopausal osteoporosis of therapy with etidronate, a diphosphonate compound that reduces bone resorption by inhibiting osteoclastic activity. The patients were randomly assigned in equal numbers to receive oral etidronate (400 mg per day) or placebo for 2 weeks, followed by a 13-week period in which no drugs were given. This sequence was repeated 10 times, for a total of 150 weeks. Daily oral supplementation with calcium and vitamin D was given throughout the study to both groups. Vertebral bone mineral content was measured by dual-photon absorptiometry; spinal radiographs were assessed to identify new vertebral fractures. Vertebral bone mineral content increased significantly (P less than 0.01) after 150 weeks of etidronate therapy (5.3 percent; 95 percent confidence interval, 2.0 to 8.6; n = 20) but decreased with placebo (-2.7 percent; 95 percent confidence interval, -7.3 to 1.9; n = 20). The difference between groups was 8.0 percentage points (P less than 0.01; 95 percent confidence interval, 2.4 to 13.6). The rates of fracture were significantly different for the period from week 60 to week 150 between the etidronate and placebo groups (6 vs. 54 fractures per 100 patient-years; P = 0.023). No adverse clinical, biochemical, or bone histomorphometric effects of treatment were observed. We conclude that at the end of nearly three years, etidronate therapy for postmenopausal osteoporosis results in significant increases in vertebral bone mineral content and, after approximately one year of treatment, a significant decrease in the rate of new vertebral fractures.