OBJECTIVE The autonomic nervous system (ANS) is a key factor in the regulation of energy balance and body fat storage;however, to what extent the physical activity during the childhood years contributes to variations in ANS function is still unclear. The present study was designed to investigate the ANS activity in lean and obese children, focusing on the differences in physical activity levels. METHODS This study was performed on 1080 school children initially recruited to the present study. In all, 24 physically active and 24 inactive obese children (> or =120% of the standard body weight) were chosen as samples. Then, 24 lean-active and 24 lean-inactive children, who were matched individually in age, gender, height, and the amount of sports activity, were carefully selected from the remaining children. METHODS Physical activity was classified as the frequency of participation in after-school sports activities (active; > or =3 times per week, inactive; nothing). The ANS activities were measured during the resting condition by means of heart rate (HR) variability power spectral analysis, which enables us to identify separate frequency components, that is, low frequency (LF; 0.03-0.15 Hz), reflecting mixed sympathetic (SNS) and parasympathetic nervous system (PNS) activity, high frequency (HF; 0.15-0.5 Hz), mainly associated with PNS activity, and total power (TP; 0.03-0.5 Hz), evaluating the overall ANS activity. The spectral powers were log transformed for statistical testing. RESULTS The lean-active group demonstrated lower resting HR as well as significantly higher TP, LF, and HF powers compared to the remaining groups. In contrast, the obese-inactive group showed significantly lower TP (P<0.05 vs the remaining groups), LF (P<0.05 vs the lean groups), and HF power (P<0.05 vs the lean groups), respectively. The obese-active and lean-inactive groups were nearly identical in all spectral parameters. The correlation analysis revealed that TP among 48 inactive children was significantly and negatively associated with the percentage of body fat (r=-0.53, P<0.001); however, such correlation among 48 active children was modest (r=-0.33, P=0.02). CONCLUSIONS Our data suggest that obese children possess reduced sympathetic as well as parasympathetic nervous activities as compared to lean children who have similar physical activity levels. Such autonomic reduction, associated with the amount of body fat in inactive state, might be an etiological factor of onset or development of childhood obesity. On the other hand, regular physical activities could contribute to enhance the overall ANS activity in both lean and obese children. These findings further imply that regular physical activity might be effective in preventing and treating obesity beginning in the childhood.