The purpose of this 7-yr prospective longitudinal study was to examine if the level and consistency of leisure-time physical activity (LTPA) during adolescence affected the quantity and distribution of lean mass (LM) and fat mass (FM) at early adulthood. The study subjects were 202 Finnish girls who were 10-13 yr old at baseline. LM and FM of the total body (TB), arms, legs, and trunk were assessed by dual-energy X-ray absorptiometry. Muscle cross-sectional area (mCSA) of the left leg was assessed by peripheral quantitative computed tomography. Scores of LTPA were obtained by questionnaire. Girls were divided into four groups comprising those with consistently low (G(LL)) or consistently high (G(HH)) physical activity, or those whose physical activity changed from low to high (G(LH)), or from high to low (G(HL)), over the 7 yr of follow-up. At baseline, no differences were found in LM, FM, and FM% among the groups in any of the body segments. By the end of the study G(HH) and G(LH) had higher values of LM of the TB, arms, legs, and trunk than that of the G(HL) and G(LL) groups (P < 0.05, respectively). High FM% of the TB was associated with low level of LTPA, but no significant differences were found in the absolute amount of FM and mCSA among the LTPA groups. Our results suggest that a consistently high level of LTPA during the transition from prepuberty to early adulthood has a positive effect on lean mass gain in girls. Participating in 5 h of LTPA per week had a significant effect on FM% but not on the absolute amount of fat mass.