During childhood, physical activity is likely the most important modifiable factor for the development of lean mass. However, the effects of normal growth and maturation must be controlled. To distinguish effects of physical activity from normal growth, longitudinal data are required. One hundred nine boys and one hundred thirteen girls, participating in the Saskatchewan Pediatric Bone Mineral Accrual Study, were repeatedly assessed for 6 yr. Age at entry was 8-15 yr. Stature, body mass, and physical activity were assessed biannually. Body composition was assessed annually by dual-energy X-ray absorptiometry. Physical activity was determined using the physical activity questionnaires for children and adolescence. Biological age was defined as years from age of peak height velocity. Data were analyzed using multilevel random-effects models. In boys, it was found that physical activity had a significant time-dependent effect on lean mass accrual of the total body (484.7+/-157.1 g), arms (69.6+/-27.2 g), legs (197.7+/-60.5 g), and trunk (249.1+/-91.4 g) (P<0.05). Although the physical activity effects were similar in the girls (total body: 306.9+/-96.6 g, arms: 31.4+/-15.5 g, legs: 162.9+/-40.0 g, and trunk: 119.6+/-58.2 g; P<0.05), boys for the same level of activity accrued, depending on the site, between 21 and 120% more absolute lean mass (g). In conclusion, habitual physical activity had a significant independent influence on the growth of lean body mass during adolescence, once biological maturity and stature were controlled.