Résumé : Dual X-ray absorptiometry (DXA) is considered a gold standard for bone measurements in the assessment of osteoporosis. Other techniques such as quantitative ultrasound (QUS) are promising to detect patients with osteoporosis-related fractures and to predict fracture risk. In this cross-sectional retrospective study, we analyzed the behavior of QUS and DXA measurements alone and in combination with regard to the presence of fractures in 320 women, 147 with nontraumatic fractures. Speed of sound (SOS), broadband ultrasound attenuation (BUA), and a third parameter derived from SOS and BUA called stiffness were measured at the calcaneus using an Achilles device (Lunar, Madison, WI). Lumbar (BMDL) and hip (BMDH( bone mineral density were measured by DXA (Hologic QDR 1000, Waltham, MA). Mean SOS, BUA, stiffness, and BMDL and BMDH were significantly lower in women with fractures compared with women without fractures. Logistic regression adjusted for age identified stiffness as the parameter most strongly associated with the presence of fracture: its sensitivity was 54% and specificity 70%. Hip BMD was second, with a sensitivity of 54% and a specificity of 69%. Combining QUS and DXA measurements did not improve the specificity nor the sensitivity. There was no difference in the odds ratios with regard to the technique that was chosen for bone assessment. In conclusion, these results suggest that low QUS measurements are associated with the presence of fractures in a way similar to DXA. In our study, the combination of QUS and DXA did not improve the discrimination of women with fractures.