The effects of voluntary isocapnic hyperpnea (VIH) training (10 h over 4 weeks, 30 min/day) on ventilatory system and running performance were studied in 15 male competitive runners, 8 of whom trained twice weekly for 3 more months. Control subjects (n = 7) performed sham-VIH. Vital capacity (VC), FEV₁, maximum voluntary ventilation (MVV), maximal inspiratory and expiratory mouth pressures, [(V)\dot]O_2max,{\dot{{V}}\hbox{O}_{{2{\rm max}}},} 4-mile run time, treadmill run time to exhaustion at 80% [(V)\dot]O_{2 max},{\dot{{V}}\hbox{O}_{{2 {\rm max}}},} serum lactate, total ventilation ([(V)\dot]_E),{(\dot{{V}}_{\rm E}),} oxygen consumption ([(V)\dot]O₂),{(\dot{{V}}\hbox{O}_{2}),} oxygen saturation and cardiac output were measured before and after 4 weeks of VIH. Respiratory parameters and 4-mile run time were measured monthly during the 3-month maintenance period. There were no significant changes in post-VIH VC and FEV₁ but MVV improved significantly (+10%). Maximal inspiratory and expiratory mouth pressures, arterial oxygen saturation and cardiac output did not change post-VIH. Respiratory and running performances were better 7- versus 1 day after VIH. Seven days post-VIH, respiratory endurance (+208%) and treadmill run time (+50%) increased significantly accompanied by significant reductions in respiratory frequency (−6%), [(V)\dot]_E{\dot{{V}}_{\rm E}} (−7%), [(V)\dot]O₂{\dot{{V}}\hbox{O}_{2}} (−6%) and lactate (−18%) during the treadmill run. Post-VIH 4-mile run time did not improve in the control group whereas it improved in the experimental group (−4%) and remained improved over a 3 month period of reduced VIH frequency. The improvements cannot be ascribed to improved blood oxygen delivery to muscle or to psychological factors.