The limited available evidence suggests that endurance training does not influence the pulmonary oxygen uptake (V(O)(2)) kinetics of pre-pubertal children. We hypothesised that, in young trained swimmers, training status-related adaptations in the V(O)(2) and heart rate (HR) kinetics would be more evident during upper body (arm cranking) than during leg cycling exercise. Eight swim-trained (T; 11.4 +/- 0.7 years) and eight untrained (UT; 11.5 +/- 0.6 years) girls completed repeated bouts of constant work rate cycling and upper body exercise at 40% of the difference between the gas exchange threshold and peak V(O)(2). The phase II V(O)(2) time constant was significantly shorter in the trained girls during upper body exercise (T: 25 +/- 3 vs. UT: 37 +/- 6 s; P < 0.01), but no training status effect was evident in the cycle response (T: 25 +/- 5 vs. UT: 25 +/- 7 s). The V(O)(2) slow component amplitude was not affected by training status or exercise modality. The time constant of the HR response was significantly faster in trained girls during both cycle (T: 31 +/- 11 vs. UT: 47 +/- 9 s; P < 0.01) and upper body (T: 33 +/- 8 vs. UT: 43 +/- 4 s; P < 0.01) exercise. The time constants of the phase II V(O)(2)and HR response were not correlated regardless of training status or exercise modality. This study demonstrates for the first time that swim-training status influences upper body V(O)(2) kinetics in pre-pubertal children, but that cycle ergometry responses are insensitive to such differences.