In menopausal women, the susceptibility to atraumatic vertebral fractures is thought to be governed by both mechanical factors (as represented by bone density) and by biologic factors such as age, body size, and dietary calcium. Whether these biologic factors independently influence fracture rates beyond the effect of bone density is a matter of controversy. To compare the relative importance of mechanical and biologic factors on vertebral fractures, we elucidated the determinants of atraumatic compression fractures in 63 menopausal women who had no chronic diseases other than osteopenia. Trabecular bone density was determined by quantitative computerized tomography of the spine. Fracture frequency was expressed as the number of compressed vertebrae per person between T5 and L4. The analysis showed that fracture frequency, bone density, and the biologic factors were closely interrelated. Fracture frequency was inversely correlated to bone density [R2 (spline model) = .40, P less than .0001], body size (r = -.26, P = .05), and dietary calcium (r = -.28, P = .04), and directly correlated with age (r = .46, P = .0002). Bone density decreased with age (r = -.65, P less than .0001), increased with body size (r = .37, P = .004), and tended to increase with dietary calcium (r = .24, P = .08). After controlling for bone density, there was no perceptible residual relationship between fracture frequency and any of these biologic factors. We conclude that mechanical factors overshadow age, body size, and dietary calcium as determinants of vertebral fracture frequency. The apparent influence of these biologic factors on fracture frequency is explained by their surrogate effects on bone density.