The purpose of this study was to determine if inspiratory muscle training (IMT) alters the oxygen cost of breathing (Vo(2RM)) during voluntary hyperpnea. Sixteen male cyclists completed 6 wk of IMT using an inspiratory load of 50% (IMT) or 15% placebo (CON) of maximal inspiratory pressure (Pi(max)). Prior to training, a maximal incremental cycle ergometer test was performed to determine Vo(2) and ventilation (V(E)) at multiple workloads. Pre- and post-training, subjects performed three separate 4-min bouts of voluntary eucapnic hyperpnea (mimic), matching V(E) that occurred at 50, 75, and 100% of Vo(2 max). Pi(max) was significantly increased (P < 0.05) by 22.5 ± 8.7% from pre- to post-IMT and remained unchanged in the CON group. The Vo(2RM) required during the mimic trial corresponded to 5.1 ± 2.5, 5.7 ± 1.4, and 11.7% ± 2.5% of the total Vo(2) (Vo(2T)) at ventilatory workloads equivalent to 50, 75, and 100% of Vo(2 max), respectively. Following IMT, the Vo(2RM) requirement significantly decreased (P < 0.05) by 1.5% (4.2 ± 1.4% of Vo(2T)) at 75% Vo(2 max) and 3.4% (8.1 ± 3.5% of Vo(2T)) at 100% Vo(2 max). No significant changes were shown in the CON group. IMT significantly reduced the O(2) cost of voluntary hyperpnea, which suggests that a reduction in the O(2) requirement of the respiratory muscles following a period of IMT may facilitate increased O(2) availability to the active muscles during exercise. These data suggest that IMT may reduce the O(2) cost of ventilation during exercise, providing an insight into mechanism(s) underpinning the reported improvements in whole body endurance performance; however, this awaits further investigation.