RATIONALE AND OBJECTIVES. The authors assessed the capability of magnetic resonance imaging in measuring bone mineral density in vivo. METHODS. Using the in-phase asymmetric spin-echo technique, the susceptibility effect (T2*suceptibility) was measured in lumbar vertebrae in vivo. Calcium density was measured with single energy quantitative computed tomography scan. The values of the susceptibility effect were correlated with calcium density. RESULTS. A statistically significant positive correlation was observed between signal loss due to susceptibility-induced inhomogeneity and the calcium density in the region of interest. The change in 1/T2*susceptibility per unit change in density was 0.114 seconds-1/mg/cm3. T2*suceptibility was 14.3 mseconds in volunteers and 17.4 mseconds in postoophorectomy patients. CONCLUSIONS. To some extent, quantification of bone density is possible using magnetic resonance imaging; however, the relatively wide dispersion of the susceptibility effect, plotted against the calcium density in the axial skeleton, indicates that other factors, such as the three-dimensional structure of bone trabeculae, may play important roles in determining magnetic resonance parameters.