We previously demonstrated that muscle-building exercise is associated with increases in serum Gla-protein, serum 1,25(OH)2D, and urinary cyclic AMP. These studies were interpreted to mean that this form of exercise increases bone formation and modifies the vitamin D-endocrine system to provide more calcium for bone. The present investigation was carried out in normal young adult white men to determine the effects of exercise on bone mineral density at weight-bearing and nonweight-bearing sites. Twelve men who had regularly engaged in muscle-building exercises (use of weights, exercise machines, or both) for at least 1 year and 50 age-matched controls (aged 19-40 years) were studied. The body weights of the two groups were not different from each other (78 +/- 2 vs. 74 +/- 1 kg, NS). Bone mineral density (BMD) of the lumbar spine, trochanter, and femoral neck was measured by dual-photon absorptiometry, and BMD of the midradius was measured by single photon absorptiometry. It was found that muscle-building exercise was associated with increased BMD at the lumbar spine (1.35 +/- 0.03 vs. 1.22 +/- 0.02 g/cm2, P less than 0.01), trochanter (0.99 +/- 0.04 vs. 0.86 +/- 0.02 g/cm2, P less than 0.01), and femoral neck (1.18 +/- 0.03 vs. 1.02 +/- 0.02 g/cm2, P less than 0.001) but not at the midradius (0.77 +/- 0.02 vs. 0.77 +/- 0.01 g/cm2, NS). These studies provide additional evidence that muscle-building exercise is associated with increases in BMD at weight-bearing sites but not at nonweight-bearing sites.