In this 3-yr, randomized, double-blind, placebo- and active-controlled study, healthy postmenopausal women with osteoporosis (55-85 yr of age) were treated with bazedoxifene 20 or 40 mg/d, raloxifene 60 mg/d, or placebo. The primary endpoint was incidence of new vertebral fractures after 36 mo; secondary endpoints included nonvertebral fractures, BMD, and bone turnover markers. Among 6847 subjects in the intent-to-treat population, the incidence of new vertebral fractures was significantly lower (p < 0.05) with bazedoxifene 20 mg (2.3%), bazedoxifene 40 mg (2.5%), and raloxifene 60 mg (2.3%) compared with placebo (4.1%), with relative risk reductions of 42%, 37%, and 42%, respectively. The treatment effect was similar among subjects with or without prevalent vertebral fracture (p = 0.89 for treatment by baseline fracture status interaction). The incidence of nonvertebral fractures with bazedoxifene or raloxifene was not significantly different from placebo. In a posthoc analysis of a subgroup of women at higher fracture risk (femoral neck T-score <or= -3.0 and/or >or=1 moderate or severe vertebral fracture or multiple mild vertebral fractures; n = 1772), bazedoxifene 20 mg showed a 50% and 44% reduction in nonvertebral fracture risk relative to placebo (p = 0.02) and raloxifene 60 mg (p = 0.05), respectively. Bazedoxifene significantly improved BMD and reduced bone marker levels (p < 0.001 versus placebo). The incidence of vasodilatation, leg cramps, and venous thromboembolic events was higher with bazedoxifene and raloxifene compared with placebo. In conclusion, bazedoxifene significantly reduced the risk of new vertebral fracture in postmenopausal women with osteoporosis and decreased the risk of nonvertebral fracture in subjects at higher fracture risk.