The purpose of this study was to investigate the relationship between cycling peak power (CPP; flywheel inertia included) and the applied braking force (F(B)) on a friction-loaded cycle ergometer in male children, adolescents, and adults. A total of 520 male subjects aged 8-20 yr performed three brief maximal sprints against three F(B): 0.245, 0.491, and 0.736 N x kg(-1) body mass (BM) (corresponding applied loads: 25 [F(B)25], 50 [F(B)50], and 75 [F(B)75] g x kg(-1) BM). For each F(B), peak power (PP) was measured (PP25, PP50 and PP75). For each subject, the highest PP was defined as CPP. Results showed that PP was dependent on F(B). In young adults PP25 underestimated CPP by more than 10%, and consequently, F(B)25 seemed to be too low for this population. However, in children, PP75 underestimated CPP by about 20%. A F(B) of 0.736 N x kg(-1) BM was definitively too high for the pediatric population. Therefore, the optimal F(B), even corrected for BM, was lower in children than in adults. The influence of growth and maturation on the force-generating capacity of the leg muscles may explain this difference. In this study, however, it was shown that the difference between PP50 and CPP was independent of age for the whole population investigated. Consequently, when flywheel inertia is included, one cycling sprint with a F(B) of 0.495 N x kg(-1) BM (corresponding applied load: 50 g x kg(-1) BM) is a feasible method for testing both children, adolescents, or young adults.