The aim of this study was to investigate the effect of respiratory muscle endurance training on endurance exercise performance in normoxic and hypoxic conditions. Eighteen healthy males were stratified for age and aerobic capacity; and randomly assigned either to the respiratory muscle endurance training (RMT = 9) or to the control training group (CON = 9). Both groups trained on a cycle-ergometer 1 h day(-1), 5 days per week for a period of 4 weeks at an intensity corresponding to 50% of peak power output. Additionally, the RMT group performed a 30-min specific endurance training of respiratory muscles (isocapnic hyperpnea) prior to the cycle ergometry. Pre, Mid, Post and 10 days after the end of training period, subjects conducted pulmonary function tests (PFTs), maximal aerobic tests in normoxia (VO(2max)NOR), and in hypoxia (VO(2max)HYPO; F(I)O(2) = 0.12); and constant-load tests at 80% of VO(2max)NOR in normoxia (CLT(NOR)), and in hypoxia (CLTHYPO). Both groups enhanced VO(2max)NOR (CON: +13.5%; RMT: +13.4%), but only the RMT group improved VO(2max)HYPO Post training (CON: -6.5%; RMT: +14.2%). Post training, the CON group increased peak power output, whereas the RMT group had higher values of maximum ventilation. Both groups increased CLT(NOR) duration (CON: +79.9%; RMT: +116.6%), but only the RMT group maintained a significantly higher CLT(NOR) 10 days after training (CON: +56.7%; RMT: +91.3%). CLT(HYPO) remained unchanged in both groups. Therefore, the respiratory muscle endurance training combined with cycle ergometer training enhanced aerobic capacity in hypoxia above the control values, but did not in normoxia. Moreover, no additional effect was obtained during constant-load exercise.