Six subjects pedaled a stationary cycle ergometer to exhaustion on three separate occasions while breathing gas mixtures of 17, 21, or 60% O2 in N2. Each subject rode for 3 min at work rates of 60, 90, 105 W, followed by 15-W increases every 3 min until exhaustion. Inspired and expired gas fractions, ventilation (V), heart rate, and blood lactate were measured. O2 uptake (VO2) and CO2 output (VCO2) were calculated for the last minute of each work rate; blood samples were drawn during the last 5 s. "Break points" for lactate, V, VCO2, V/VO2, and expired oxygen fraction (FEO2) were mathematically determined. VO2 was not significantly different at any work rate among the three different conditions. Nor did maximal VO2 differ significantly among the three treatments (P greater than 0.05). Lactate concentrations were significantly lower during hyperoxia and significantly higher during hypoxia compared with normoxia. Lactate values at exhaustion were not significantly different among the three treatments. Four subjects were able to work for a longer period of time during hyperoxic breathing. The variations in lactate accumulation as reported in this study cannot be explained on the basis of differences in VO2. The results of this research lend support to the hypothesis that differences in the performance of subjects breathing altered fractions of inspired oxygen may be caused by differences in lactate (or H+) accumulation.