The aim of the present research was to determine the association between historical physical activity and baseline bone measurements in a group of 223 postmenopausal women participating in a clinical trial in Pittsburgh, Pennsylvania, from 1981 to 1986 by evaluating the effect of moderate physical activity on bone loss. Historical physical activity was assessed by a survey which divided the life span into four time periods (14-21, 22-34, 35-50, and 50+ years) and inquired about participation in leisure time physical activities for each period. From the responses, kilocalories of energy expenditure were calculated. Cortical bone density and area were measured in the radius with a computerized tomography scanner. The historical physical activity survey was administered a second time two to three months after the initial test to a 10% random sample of the women in order to determine the test-retest reliability of the instrument. Since the measurements of historical physical activity proved to be reliable, estimates of kilocalories determined for the entire population of women were correlated with bone area and density. A significant relation was found to exist between historical physical activity and dimensions of adult bone, particularly bone area. This association remained significant after adjustment for potential confounding variables and seemed to be strongest in the earlier age periods. To the authors' knowledge, this is the first report of a significant association between historical physical activity and bone.