The mechanisms of exertional dyspnea relief in response to supplemental oxygen (O2) in chronic airflow limitation (CAL) are not precisely known and are likely multifactorial. To explore factors contributing to the relief of dyspnea after oxygen administration, 11 patients with severe CAL (FEV1.0 = 39 +/- 3% predicted, mean +/- SEM) and mild hypoxemia (resting PaO2 = 74 +/- 2 mm Hg) breathed room air (RA) and 60% O2 during exercise at approximately 50% of their maximal incremental exercise capacity. Breathlessness ratings (Borg scale), endurance time, respiratory drive (change in mouth occlusion pressure over the first 0.1 s of inspiration, P0.1), ventilation (VE), breathing pattern, operational lung volumes, gas exchange, and metabolic parameters were compared during RA and 60% O2. PaO2 at exercise cessation during RA and 60% O2 was 65 +/- 3 mm Hg and 226 +/- 12 mm Hg, respectively (p < 0.001). With 60% O2, the mean of individual Borg/time slopes fell significantly (p < 0.05) by 23 +/- 12% and was associated with a 35 +/- 11% increase (p < 0.01) in endurance time (r = -0.64, p < 0.05). During 60% O2, slopes of P0.1 and lactate over time also fell significantly (p < 0.05), whereas delta PaCO2/time did not change significantly. At a standardized time near end-exercise, Borg, VE, and P0.1 changed during 60% O2 by -0.8 +/- 0.3 (p < 0.05), -4.1 +/- 2.0 L/min (p = 0.07), and -1.3 +/- 0.5 cm H2O/s (p < 0.05), respectively. Slopes of Borg/VE, Borg/lactate, and VE/lactate were essentially superimposable during tests on RA and O2: Borg, lactate, and VE all fell proportionally during hyperoxia. In patients with CAL and mild exercise hypoxemia, relief of exertional breathlessness during hyperoxia is explained by reduced ventilatory demand in association with reduced blood lactate levels.