To assess the effects of respiratory muscle training (RMT) on maximum oxygen uptake (V'O_2max) in normoxia and hypoxia, 9 healthy males (age 24±4 years; stature 1.75±0.08 m; body mass 72±9 kg; mean±SD) performed on different days maximal incremental tests on a cycle ergometer in normoxia and normobaric hypoxia (FIO₂ = 0.11), before and after 8 weeks of RMT (5 days/week). During each test, gas exchange variables were measured breath-by-breath by a metabolimeter. After RMT, no changes in cardiorespiratory and metabolic variables were detected at maximal exercise in normoxia. On the contrary, in hypoxia expired and alveolar ventilation (V'E and V'A, respectively) at maximal exercise were significantly higher than pre-training condition (+12 and +13%, respectively; P<0.05). Accordingly, alveolar O₂ partial pressure (PAO₂) after RMT significantly increased by ~10%. Nevertheless, arterial PO₂ and V'O_2max did not change with respect to pre-training condition. In conclusion, RMT improved respiratory function but did not have any effect on V'O_2max, neither under normoxic nor hypoxic condition. In hypoxia, the significant increase in V'E and V'A at maximum exercise after training lead to higher alveolar but not arterial PO₂ values, revealing an increased A-a gradient. This result, according to the theoretical models of V'O_2max limitation, seems to contradict the lack of V'O_2max increase in hypoxia, suggesting a possible role of increased ventilation-perfusion mismatch.