Quantitative ultrasound (US) measurements have been shown to be a new technique assessing bone status. This study aimed to assess a new US instrument, the DBM Sonic 1200(R) (IGEA) which permits the measurement of the speed of sound in the proximal phalanges (SOSp) of the hand. The results obtained were compared with DXA (SOPHOS) and US measurements at the calcaneus (Achilles(R) LUNAR). The in vivo precision expressed by coefficient of variation was 0.91%. Ultrasound measurements of phalanges were significantly correlated with BMD in the entire group of 90 subjects: osteoporotic patients (n = 47) and controls (n = 43) (r = 0.44, femoral neck and 0.45, lumbar spine, P < 0.01). A significant correlation was also found in the control group (r = 0.33, lumbar spine and 0.38, femoral neck, P < 0.05) but not in the osteoporotic group (r = 0.3, lumbar spine and 0.17, femoral neck, P > 0.05). Mean values for 31 postmenopausal, osteoporotic women and age-matched controls showed a significant decrease in US measurements at the phalanges (P < 0.05) and the calcaneus (P < 0.01) as well as bone mineral density (BMD) at the spine and femoral neck (P < 0.01) in the osteoporotic group. A decision threshold for a sensitivity of 80% for osteoporotic fractures resulted in a specificity value of only 37% for SOSp, between 53 to 65% for calcaneus US measurements and 45 to 56% for BMD. The Z score, the odds ratio, the ROC curves, and areas under the curves plotted for the subgroup of 31 fractures and their healthy controls showed poorer values for SOSp than BMD and calcaneus US measurements. In conclusion, US measurements of phalanges seem to be less efficient than calcaneus US and BMD measurements to distinguish osteoporotic from healthy women. Other studies and also prospective studies are required to assess the interest in fracture risk assessment.