Exercise-induced changes in QRS voltage have been used to evaluate left ventricular function in adults, children, and adolescents. The purpose of this study was to investigate changes in QRS voltage with exercise in a population of healthy children and adolescents. Eighty-two normal subjects (61 boys and 21 girls) ages 4-18 yr (mean = 12 yr) had maximal exercise testing performed according to Bruce treadmill protocol. Subjects were exercised to exhaustion with continuous monitoring of a VF, V2, and V5. Twelve lead electrocardiograms were recorded at each stage of exercise. Maximum heart rate obtained was 180-212 (mean = 197). Resting and exercise QRS voltages were measured in a VF and V5 averaged over five consecutive QRS complexes. From 12 lead electrocardiographic recordings, frontal and horizontal QRS vectors were calculated. There was no change in the frontal plane QRS vector with exercise (P = 0.4). There was a posterior change in horizontal vector at peak exercise (P less than 0.001) with a mean shift of 14.5 degrees. The R wave voltage diminished in V5 (P less than 0.001, mean decrease of 0.39 mV). However, V5 R wave voltage was increased or unchanged in 23 of the 61 boys and in three of the 21 girls. In males 10 yr of age or younger, 72% had no change or increased R wave voltage compared with 21% for males older than 10 yr (P less than 0.05). These findings are clearly different from the QRS changes with exercise in healthy adults.