Abstract Introduction Bazedoxifene has been shown to significantly decrease the risk of vertebral fractures in postmenopausal women. No significant effect was noted on the risk of clinical fractures, but fracture risk reduction was reported in a post hoc subgroup analysis in a high risk group categorised on the basis of BMD and prior fracture. Aims The aim of this study was to re-evaluate the efficacy of bazedoxifene on fracture outcomes avoiding subgroup analysis by examining the efficacy of intervention as a function of fracture risk. Methods The phase III study was a double-blind, randomised, placebo- and raloxifene-controlled randomised 3-year multinational study that enrolled 7492 osteoporotic women aged 55 years or more (mean age = 66 years). For the present analysis, women taking raloxifene were excluded ( n = 1849), and we compared the effects of two doses of bazedoxifene (20 and 40 mg daily combined) with placebo on the risk of all clinical fractures as well as the risk of morphometric vertebral fracture. The risk of a major osteoporotic fracture was assessed using region specific FRAX ® algorithms, and the relationship between pre hoc 10-year fracture probabilities and efficacy examined by Poisson regression. Results Overall, bazedoxifene was associated with a significant 39% decrease in incident morphometric vertebral fractures (hazard ratio HR = 0.61; 95% CI = 0.43–0.86; p = 0.005) and a non-statistically significant 16% decrease in all clinical fractures (hazard ratio HR = 0.84; 95% CI = 0.67–1.06; p = 0.14) compared to placebo. Hazard ratios for the effect of bazedoxifene on all clinical fractures decreased with increasing fracture probability. In patients with 10-year fracture probabilities at or above 16%, bazedoxifene was associated with a significant decrease in the risk of all clinical fractures. The 16% probability threshold corresponded to the 80th percentile of the study population. Hazard ratios for the effect of bazedoxifene on morphometric vertebral fractures also decreased with increasing fracture probability. In patients with 10-year fracture probabilities above 6.9% (corresponding to the 41st percentile), bazedoxifene was associated with a significant decrease in the risk of morphometric vertebral fractures. At equivalent fracture probability percentiles, the treatment effect of bazedoxifene was greater on vertebral fracture risk than on the risk of all clinical fractures. For example, at the 90th percentile of FRAX ® probability, bazedoxifene was associated with a relative risk reduction of 33% (95% CI = 7–51%) for all clinical fractures and 51% reduction (95% CI = 21–69%) for morphometric vertebral fractures. The findings were robust to several sensitivity analyses. Conclusion Bazedoxifene (20 and 40 mg doses combined) significantly decreased the risk of all clinical fractures and morphometric vertebral fractures in women at or above a FRAX ® based fracture probability threshold. These results, consistent with the previous subgroup analysis, suggest that bazedoxifene should be targeted preferentially to women at high fracture risk.