Previous studies on the relation between moderate physical activity and bone mass have observed conflicting results. Many of these studies have not dissociated the role of physical activity by age groups and in relation to the period of peak bone mass formation. Our cross-sectional analysis of the baseline data of a longitudinal study of 273 women aged 21-40 attempted to evaluate the role of moderate physical activity on bone mass around the period of peak bone mass attainment. The analyses were carried out separately for the two age groups--21-30 and 31-40--and had also taken into account the effects of age, dietary calcium intake, and lean body mass on bone mineral density (BMD). The total metabolic equivalent values (MET) of leisure time physical activity was based on the MET values for each activity and the reported time spent on each activity in the past year. The results indicated that among the younger group of women, high level of leisure time physical activity was associated with higher bone mass at both the spine and the hip. Additive effects of physical activity and dietary calcium intake on the spine and the hip BMD were observed. Together with age and lean body mass, physical activity and dietary calcium intake accounted for 19% of the variances of bone mineral at the spine and 9-11% at the hip. Among women aged 31-40, presumably after the peak bone mass formation, lean body mass as well as fat mass have independent strong association with BMD. Physical activity was not associated with bone mass in this age group.