The aim of the present study was to investigate the effect of a long-term intensive swimming training program on the aerobic potential of prepubertal girls and to show that the choice of the laboratory test is a very important factor in order to point out the physiological adaptations linked to the training regimen. The subjects consisted in 14 prepubertal girls, five of whom came from a swimming club (GS) and nine, af the same age (9 years old), constituted the control group (CG). The girls took part in three experiments in which the aim was to determine the maximal oxygen uptake. They were organised in several steps (5-7 depending on the subject) of 2 min each against increasing forces until exhaustion. Several parameters, including oxygen uptake (VO 2 ), carbon dioxyde output (VCO 2 ), minute ventilation (VE), heart rate (HR) and exchange ratio (R), were recorded in the last 30 s of each step. The 2 first investigations were performed on a swim bench (peak VO 2 ) at the beginning of the training program and one year later. The third tests was realised on a bicycle ergometer (VO 2 max) 5 months after the second swim bench test. During the training period, the girls swimmers swam 10-12 h per week (1h-1h30 twice a day; 5 days per week) and performed 15-20 km a week. There was no significant difference between the two groups at the beginning of the training program whatever the physiological parameter (peak VO 2 : GS: 26 +/- 4 mL.min - 1 kg - 1 , CG: 25 +/- 3 mL.min - 1 kg- 1 NS; HR: GS: 167 +/- 10 beat.min- 1 , CG: 170 +/- 14 beat.min - 1 , NS). However, after one year of training, the peak VO2 of the girls swimmers were much higher (+36 %) than those of the control group (GS: 34 +/- 5 mL.min - 1 kg- 1 , CG: 25 +/- 4 mL.min - 1 kg- 1 , P <0.01). Similar results were obtained regarding maximal aerobic power (Pmax: GS: 44 +/- 10 watts, CG: 23 +/- 04 watts, P < 0.001). There was however no significant difference for the other parameters (HR: GS : 184 beat.min-1, CG: 180 beat.min - 1 , NS). After one year and a half of training, VO 2 max was however only a little beat much higher for the girls swimmers than for the control group (GS: 46 +/- 01 mL.min - 1 kg - 1 , CG: 41 +/- 4 mL.min - 1 kg - 1 , P < 0.05). This was also true for Pmax (Pmax: GS: 130 +/- 15 watts, CG: 102 +/- 23 watts, P < 0.05). No significant difference appeared for the other criteria (HR: GS: 208 +/- 03 beat.min - 1 , CG: 204 +/- 05 beat.min- 1 , NS). For the both groups, peak VO 2 was lower than VO 2 max (GN: P < 0.01; GC: P <0.001). These results show, in one hand, that physiological