Recently, an ultrasound (US) device for measurement of amplitude-dependent speed of sound in four proximal phalanges of the hand (DBM Sonic 1200, IGEA, Carpi, Mo, Italy) has been introduced but has not been thoroughly investigated in populations at most risk for fragility fractures (i.e., elderly women). As part of the Malmö Osteoporosis Prospective Risk Assessment study (OPRA), we investigated 1044 randomly selected women, all 75 yr of age, with US of the phalanges and, for comparison, also with two more established methods for bone mass measurement: US of the calcaneus and dual-energy X-ray absorptiometry (DXA) of the hip and spine, both methods having an ability to predict fracture. A self-assessment questionnaire was used to obtain information on previous fracture and age at fracture event. We found a low correlation between US of the phalanges and US of the calcaneus speed of sound (SoS) (r = 0.11, p < 0.01), US of the calcaneus (stiffness) (r = 0.09, p < 0.05), DXA of the femoral neck (r = 0.09, p < 0.05), and DXA of the spine (r = 0.10, p < 0.01) and no significant correlation between US of the phalanges and US of the calcaneus broadband ultrasound attenuation (BUA) and DXA trochanter. Also, no differences in US of the phalanges were found when comparing women without any fracture with women with at least one fracture, whereas US of the calcaneus (SoS, BUA, and stiffness) and DXA of the femoral neck, trochanter, and spine were all lower in the women with a fracture history (p < 0.0001). In addition, the precision of the US of the phalanges method was evaluated and found to be lower in these elderly women, compared to the precision reported by others and the manufacturer. In summary, the present data indicate that US of the phalanges is not a usable tool for estimating fracture risk in an elderly female population.