This project studies the effects of exercise on bones, focusing on the two most important processes by which bones respond to loading: (1) growth (modeling) in which a bone increases its size, and (2) repair (Haversian remodeling) in which bones replace old, damaged tissue. Preliminary data indicate that the relative percentage of growth versus repair tends to vary within individuals inversely down the limbs, with more growth occurring near the body's center and more remodeling occurring distally. It is proposed that growth preferentially occurs where there is a low cost to additional mass increase, but that repair is favored in regions such as the ends of the limb where mass increases impose an exponentially higher energetic cost and which, therefore, tend to be thinner and weaker. Modulation of growth versus repair responses to loading suggest that different regions of the skeleton are adapted to tolerate different strains levels. This proposal has significance for anthropological, palaeontological and other osteologcial research on humans and other vertebrates. Of special interest is testing the hypothesis that recent, industrial-era humans tend to have overall thinner bones than hunter-gatherers and early farmers because of generally low levels of mechanical loading, particularly during childhood.