Results of trials on the effects of physical activity on bone mineral density (BMD) are conflicting. The current hypothesis was that calcium intake modifies the bone response to physical activity. Published trials on physical activity and bone density were reviewed, and the results of 17 trials are summarized. Physical activity has beneficial effects on BMD at high calcium intakes, with no effect at mean calcium intakes less than a mean of 1000 mg/day. The modifying effect of calcium intake on BMD among exercise groups is more pronounced in the lumbar spine than in the radius. This analysis may explain conflicting results of trials on physical activity and calcium effects on bone. Controlled trials designed to test adequately this hypothesis are needed.