The purpose of this study was to cross-validate the age-specific Freedson MET equation among children and adolescents of varying body size. Sixty-seven children (41 boys; 26 girls) between 6 and 13 years completed five 3-min trials (1.6, 3.2, 4.0, 4.8, and 6.4 km h(-1)) on a motorized treadmill. During each trial, participants wore an Actigraph accelerometer while oxygen consumption was assessed by indirect calorimetry. Using the Actigraph activity counts, predicted MET values were determined with the age-specific Freedson equation and were compared with measured MET values using dependent t tests. Participants were divided into body size categories based on their calculated body surface area (BSA, m2) (small: BSA<or=0.96 m2; medium: 0.96 m2<BSA<or=1.20 m2; large: BSA>1.20 m2) to determine if body size influenced the difference between measured and predicted MET values. The measured MET value was similar to the predicted MET value at the slowest treadmill speed (1.6 km h(-1)) (2.3 vs. 2.3 METs); however, the measured MET value was lower than the predicted MET value at the remaining speeds (3.2, 4.0, 4.8, and 6.4 km h(-1)) (P<0.001). With the exception of the fastest treadmill speed (6.4 km h(-1)), the mean difference between the measured and predicted MET values was greater between the two smaller BSA categories compared to the largest BSA category. The results suggest that the age-specific Freedson child equation significantly overestimates energy expenditure (METs) during locomotor speeds between 3.2 and 6.4 km h(-1). This effect was primary observed among relatively smaller children.