OBJECTIVE: Significant changes in body composition occur during lifetime. This longitudinal study (8.0 +/- 0.8 yrs) in a cohort of healthy sedentary and physically active men (n = 78) and women (n = 53), aged 20 to 74 yr describes: 1) the longitudinal changes in weight and body composition and 2) their associations with age and physical activity. Method: Fat-free mass (FFM) and body fat (BF) were assessed by bioelectrical impedance analysis (BIA). Subjects who regularly performed >3 hours per week of endurance type physical activity were classified as "Active". Others were classified as "Sedentary". Subjects were also separated by age (<45 yr vs > or =45 yr). RESULTS: FFM increased by 1.7 +/- 2.8 kg in men <45 yr who gained 4.0 +/- 5.0 kg of body weight and was maintained (0.5 +/- 1.6 kg) in women <45 y who gained 1.6 +/- 3.0 kg of weight. A weight gain of 1.2 +/- 3.3 kg in men > or =45 yr was accompanied by stable FFM (-0.1 +/- 2.3 kg), and of 1.0 +/- 3.2 kg was accompanied by a loss of FFM in women > or =45 yr. In active men > or =45 yr, maintenance of FFM was associated with smaller weight gains than in sedentary; sedentary men > or =45 yr decreased FFM with larger weight gains than active subjects. Sedentary women <45 yr were able to gain FFM; the active women maintained, but did not gain FFM with smaller weight gains than in sedentary women. FFM decreased in >/=45 yr women despite of small weight gains. CONCLUSION: Weight change is clearly associated with a change in FFM. Weight gain is necessary to offset age-related FFM loss between 20 and 74 yrs. In active men, a FFM increase was associated with less weight gain than sedentary men. Future studies should evaluate the threshold of weight change and the level of physical activity necessary to prevent age-related losses of FFM.