Training-induced adaptations in aerobic fitness have been extensively studied in adults, and some exercise scientists have recommended similar training programmes for young people. However, the subject of the response to aerobic training of children and adolescents is controversial. The effects of exercise training on prepubertal children are particularly debatable. The latter may be partly explained by different training designs, which make comparisons between studies very problematic. We have analysed the procedures applied to protocol design and training methods to highlight the real impact of aerobic training on the peak oxygen uptake (V-dotO2) of healthy children and adolescents. In accordance with previously published reviews on trainability in youngsters, research papers were rejected from the final analysis according to criteria such as the lack of a control group, an unclear training protocol, inappropriate statistical procedures, small sample size, studies with trained or special populations, or with no peak V-dotO2 data. Factors such as maturity, group constitution, consistency between training and testing procedures, drop out rates, or attendance were considered, and possible associations with changes in peak V-dotO2 with training are discussed. From 51 studies reviewed, 22 were finally retained. In most of the studies, there was a considerable lack of research regarding circumpubertal individuals in general, and particularly in girls. The results suggest that methodologically listed parameters will exert a potential influence on the magnitude of peak V-dotO2 improvement. Even if little difference is reported for each parameter, it is suggested that the sum of errors will result in a significant bias in the assessment of training effects. The characteristics of each training protocol were also analysed to establish their respective potential influence on peak V-dotO2 changes. In general, aerobic training leads to a mean improvement of 5-6% in the peak V-dotO2 of children or adolescents. When only studies that reported significant training effect were taken into account, the mean improvement in peak V-dotO2 rose to 8-10%. Results suggested that intensities higher than 80% of maximal heart rate are necessary to expect a significant improvement in peak V-dotO2. There is clearly a need for longitudinal or cross-sectional studies that investigate the relationship between maturity and training with carefully monitored programmes. Further research is also needed to compare interval training and continuous training.