Background: Obesity is a major public health problem in the United States. The role of physical activity and formal exercise in controlling body weight has not been clearly determined. Objective: This study determined the magnitude of change in body weight and composition across sex, race, and age in response to 20 wk of endurance training. Design: Men and women (n = 557) of various ages (16-65 y) and 2 races (black and white) exercised on cycle ergometers 3 d/wk for a total of 60 exercise sessions starting at 55% of maximal oxygen consumption ( · VO 2 max) for 30 min/session and building to 75% of · VO 2 max for 50 min/session, where it was maintatined during the last 6 wk. Skinfold-thickness measurements, circumferences, body composition (by hydrostatic weighing), and body fat distribution (by computed tomography scan at L4-L5 and the waist-hip ratio) were determined before and after training. Results: All skinfold-thickness and circumference measures, waist-hip ratio, body mass index, total body mass, fat mass, percentage body fat, and computed tomography scan measures of total, subcutaneous, and visceral abdominal fat decreased with training, whereas total body density and fat-free mass increased. These changes were significant, but small. There were several differences in training response by sex and race, but not by age. Conclusions: A short-term exercise intervention can induce favorable changes in body composition, but the magnitude of these changes is of limited biological significance. Increasing physical activity likely has a major effect on body-composition and fat distribution characteristics only when it is of a greater magnitude and sustained for much longer periods Am J Clin Nutr 1999;70:346-52.