We assessed the rates of vertebral fracture in patients with postmenopausal osteoporosis. Forty-five patients were not treated (91 person-years of observation); 59 were treated conventionally, with calcium (alone or combined with estrogen) or vitamin D or both (218 years); and 61 were treated with sodium fluoride combined with conventional therapy (251 years). The fracture rate (per thousand person-years) was 834 in untreated patients, 419 in those given calcium with or without vitamin D, 304 in those given fluoride and calcium with or without vitamin D, 181 in those given estrogen and calcium with or without vitamin D, and 53 in those given fluoride, estrogen, and calcium with or without vitamin D. It was reduced in all treatment groups (P less than 0.001 for calcium and P less than 1 x 10(-6) for other combinations); fluoride (one years of treatment) and estrogen (but not vitamin D) independently reduced the rate from that observed with calcium alone (P less than 0.001). The combination of calcium fluoride, and estrogen was more effective than any other combination (P less than 0.001). These results provide grounds for optimism about the efficacy of combinations of available agents with sodium fluoride for fracture in postmenopausal osteoporosis.