AIM: To investigate if training during hypoxia (H) improves the adaptation of muscle oxidative function compared with normoxic (N) training performed at the same relative intensity. METHOD: Eight untrained volunteers performed one-legged cycle training during 4 weeks in a low-pressure chamber. One leg was trained under N conditions and the other leg under hypobaric hypoxia (526 mmHg) at the same relative intensity as during N (65% of maximal power output, W(max)). Muscle biopsies were taken from vastus lateralis before and after the training period. Muscle samples were analysed for the activities of oxidative enzymes [citrate synthase (CS) and cytochrome c oxidase (COX)] and mitochondrial respiratory function. RESULTS: W(max) increased with more than 30% over the training period during both N and H. CS activity increased significantly after training during N conditions (+20.8%, P < 0.05) but remained unchanged after H training (+4.5%, ns) with a significant difference between conditions (P < 0.05 H vs. N). COX activity was not significantly changed by training and was not different between exercise conditions [+14.6 (N) vs. -2.3% (H), ns]. Maximal ADP stimulated respiration (state 3) expressed per weight of muscle tended to increase after N (+31.2%, P < 0.08) but not after H training (+3.2%, ns). No changes were found in state four respiration, respiratory control index, P/O ratio, mitochondrial Ca(2+) resistance and apparent Km for oxygen. CONCLUSION: The training-induced increase in muscle oxidative function observed during N was abolished during H. Altitude training may thus be disadvantageous for adaptation of muscle oxidative function.