Fifty-eight children and adolescents of both sexes, aged 8-16, were tested on a treadmill using two different protocols. The well-known Bruce-protocol has the disadvantages of steep incline and large increments at each step. A new protocol (Oslo-protocol) with less incline and smaller increments was compared to the Bruce-protocol. The results from the two protocols showed no differences with regard to peak oxygen uptake (VO2peak) or peak heart rate (HRpeak). However, the respiratory exchange ratio (R) and blood lactate concentration [La-] showed higher values when the Bruce-protocol was used. The study also indicated that the often used criteria of HRpeak, R and achievement of a plateau in VO2 to estimate VO2peak, were not reliable indicators in either protocol. When time to exhaustion was used as an estimation of aerobic endurance level, the Oslo-protocol discriminated better than the Bruce-protocol. As a conclusion, the results indicate that none of the criteria may be used as a reliable indicator of having achieved VO2peak. An experienced testleader may be essential to define when VO2peak has been reached in children. On the basis of the results from the current study, the Oslo-protocol seems suitable as a test-protocol when testing children and adolescents for VO2peak.