Quantitative ultrasound (QUS) of the heel is becoming increasingly popular for the assessment of skeletal status, although there appears to be a general lack of agreement regarding which side to measure. The purpose of the present study was to evaluate possible side differences (right versus left) in heel QUS within the general population, including children (10-15 years old, n=406), adults (26-33 years old, n=339), and elderly subjects (60-75 years old, n=455) of both genders (818 females and 382 males), and to examine the impact of these differences on prevalence estimates of osteoporosis and individual fracture risk assessment. All participants had both their heels measured twice with the Sahara device, which measures broadband ultrasound attenuation (BUA) and speed of sound (SOS) through the os calcis; a composite parameter, that is, quantitative ultrasound index (QUI) and an estimate of heel BMD (eBMD) were also derived. Significant side differences were detected for BUA and SOS (P<0.05), but not for QUI or eBMD. Contralateral differences were rather small in absolute terms, but were in the order of 12.6% for BUA, 0.72% for SOS, 7.9% for QUI, and 9.9% for eBMD, when expressed as percentage of the mean values for the two heels. Bilateral differences appeared to vary across age and gender. Significant correlations between QUS indices of the right and left heel were observed (r=0.75-0.85; P<0.001), which seemed to be stronger among the elderly and among male individuals. Prevalence rates of osteopenia and osteoporosis were not significantly different when estimated from eBMD T-scores for the one or the other foot (chi(2)=1.781, df=2, P=0.410). However, cross-classification analysis revealed that only 84% of the subjects classified into each risk category by the two calcanei were actually the same persons. In conclusion, results from the present study strongly suggest that QUS measurements of opposite heels may not be equivalent with respect to the evaluation of bone status and classification of individual fracture risk assessment, although the degree of discrepancy appears to be related to the primary outcome of interest.