BACKGROUND: The influence of physical activity on body weight in children and adolescents is controversial. OBJECTIVE: The objective was to test the hypothesis that the intensity and duration of physical activity differ between obese and normal-weight adolescents, with no difference in estimated energy expenditure. DESIGN: We compared physical activity in 18 (8 males, 10 females) obese [body mass index (in kg/m(2)) > 30] adolescents (14-19 y) with that in a matched, normal-weight (BMI < 27) control group. Total energy expenditure (TEE) was measured with the doubly labeled water method, and physical activity was measured simultaneously by accelerometry. The physical activity level was determined as the ratio of TEE to the resting metabolic rate (RMR) and activity energy expenditure as 0.9 TEE minus RMR. Accelerometry data included total physical activity (counts x min(-1) x d(-1)), accumulated and continuous duration of activity, and continuous 10-min periods of physical activity of moderate intensity. RESULTS: There was no significant difference in adjusted (analysis of covariance) TEE, RMR, or AEE between groups. The physical activity level was significantly lower (P < 0.05) in the obese group. No sex x group interaction was observed. Differences in total physical activity (P < 0.001), accumulated time (P < 0.05), continuous time (P < 0.01), and continuous 10-min periods of physical activity of moderate intensity (P < 0.01) were observed between groups. CONCLUSIONS: Obese adolescents are less physically active than are normal-weight adolescents, but physical activity-related energy expenditure is not significantly different between groups. The data suggest that physical activity is not necessarily equivalent to the energy costs of activity.