We utilized non-invasive methods and novel computational approaches to examine the effects of acutely varying fractional inspired O2 (FIO2) on convective and diffusive steps of O2 transport and muscle tissue (de)oxygenation during incremental cycling to exhaustion in 10 Tier 3 and 4 endurance athletes breathing either 0.152, 0.209, or 0.298 FIO2. At submaximal work rates (100-275 W) in hypoxia, higher cardiac output compensated for lower arterial O2 content. At maximal work rate, convective O2 transport was lower in hypoxia (mean [95%CI]: 5.37 [5.14-5.59] L/min, q textless 0.0001) and higher in hyperoxia (6.84 [6.50-7.18] L/min, q = 0.043) compared to normoxia (6.56 [6.16-6.95] L/min), whereas O2 diffusive conductance did not differ between conditions (94 [82-106], 98 [83-112], 98[87-109] mL/min/mmHg for hypoxia, normoxia and hyperoxia, respectively, p = 0.490). Consequently, maximal O2 uptake (V̇O2max) was lower in hypoxia (4.06 [3.85-4.27] L/min, q textless 0.0001) and higher in hyperoxia (5.02 [4.85-5.19] L/min, q = 0.003) compared to normoxia (4.83 [4.63-5.02] L/min). In hypoxia, muscle tissue saturation index was 1%-4%-units lower compared to normoxia and hyperoxia during submaximal cycling but similar at maximal work rate. In summary, central and peripheral compensatory mechanisms maintained O2 uptake despite altered FIO2 at submaximal work rates. At maximal work rate the effects of hypoxia and hyperoxia on V̇O2max were mediated through convective O2 transport.