Pulmonary O2 uptake (V(O₂p)) and leg blood flow (LBF) kinetics were examined at the onset of moderate-intensity exercise, during hyperventilation with and without associated hypocapnic alkalosis. Seven male subjects (25 ± 6 years old; mean ± SD) performed alternate-leg knee-extension exercise from baseline to moderate-intensity exercise (80% of estimated lactate threshold) and completed four to six repetitions for each of the following three conditions: (i) control [CON; end-tidal partial pressure of CO2 (P(ET, CO₂)) ~40 mmHg], i.e. normal breathing with normal inspired CO2 (0.03%); (ii) hypocapnia (HYPO; P(ET, CO₂) ~20 mmHg), i.e. sustained hyperventilation with normal inspired CO2 (0.03%); and (iii) normocapnia (NORMO; P(ET, CO₂) ~40 mmHg), i.e. sustained hyperventilation with elevated inspired CO2 (~5%). The V(O₂p) was measured breath by breath using mass spectrometry and a volume turbine. Femoral artery mean blood velocity was measured by Doppler ultrasound, and LBF was calculated from femoral artery diameter and mean blood velocity. Phase 2 V(O₂p) kinetics (τV(O₂p)) was different (P < 0.05) amongst all three conditions (CON, 19 ± 7 s; HYPO, 43 ± 17 s; and NORMO, 30 ± 8 s), while LBF kinetics (τLBF) was slower (P < 0.05) in HYPO (31 ± 9 s) compared with both CON (19 ± 3 s) and NORMO (20 ± 6 s). Similar to previous findings, HYPO was associated with slower V(O₂p) and LBF kinetics compared with CON. In the present study, preventing the fall in end-tidal P(CO₂) (NORMO) restored LBF kinetics, but not V(O₂p) kinetics, which remained 'slowed' relative to CON. These data suggest that the hyperventilation manoeuvre itself (i.e. independent of induced hypocapnic alkalosis) may contribute to the slower V(O₂p) kinetics observed during HYPO.