Perturbations to the intracellular milieu during muscular work can include the accumulation of fatigue‐inducing metabolites in skeletal muscle. It has been suggested that oxygen availability contributes to fatigue by limiting the bioenergetic pathways that support contractile activity. Our purpose was to examine the role of intracellular oxygen (PO2) in the disruption of muscle bioenergetics and consequent fatigue in vivo. We hypothesized that fatigue (decrease in peak power) would be associated with acidosis and inorganic phosphate (Pi), but not with PO2, which would remain above the critical threshold for limiting mitochondrial respiration. Seven untrained adults (35 ± 3 yr, mean ± SD; 5 F) performed 120 maximal, isotonic contractions (1 every 2 s) at 20% of peak torque on an MR‐compatible knee extension ergometer in a 3‐tesla magnetic resonance (MR) system. Phosphorus (2000‐ms repetition time) and proton (30‐ms repetition time) MR spectra were collected concurrently with a dual‐tuned (1H/31P) surface coil on the vastus lateralis muscle during 60 s of rest and 4 min of contractions; spectra were averaged to obtain 20‐s temporal resolution. Peaks corresponding to phosphocreatine (PCr), Pi, ATP and deoxymyoglobin (dMb) were fit using jMRUI. [PCr] and [Pi] were calculated assuming ATP = 8.2 mM at rest. Oxidative ATP synthesis (ATPOX, mM·s‐1) was calculated as Vmax·(1+K·([ADP][Pi][ATP]‐1))‐1 where Vmax was determined from PCr resynthesis kinetics, [ADP] was calculated using the creatine kinase equilibrium, and K = 1.1 mM. PO2 was calculated using the oxygen binding curve for myoglobin (Mb; P50 · [(1‐□)/□]), where □ is fractional Mb desaturation relative to maximal dMb during cuff occlusion, assuming a P50 for Mb of 2.39 Torr. Statistics: paired t‐tests and Pearson correlations on individual data. PO2 was fit using bilinear regression to obtain the time to PO2 plateau. Peak power declined to 63.1 ± 16.2% of initial (p < 0.01), indicating significant fatigue. PCr declined from 31.6 ± 1.7 mM at rest to 5.6 ± 4.4 mM, Pi increased from 2.9 ± 0.4 mM to 22.0 ± 3.1 mM, and pH declined to 6.61 ± 0.17 (p < 0.01, all). PO2was 36.5 ± 6.4 Torr at rest and decreased rapidly at the onset of contractions. By 27.4 ± 2.4 s, PO2 plateaued at 1.54 ± 0.6 Torr, well above the critical threshold of 0.35 Torr. ATPOX increased steadily from the start of contractions to 0.49 ± 0.16 mM·s‐1 at 100 s, which is well beyond the time when PO2 plateaued, further indicating that PO2 did not limit oxidative phosphorylation. Changes in pH and Pi were linearly associated with changes in peak power (r = 0.74 ± 0.2 and r = ‐0.66 ± 0.1, respectively), consistent with their role in myosin cross‐bridge impairment in fatigue. In contrast, there was no linear relationship between PO2 and fatigue, pH, or [Pi] during or at the end of the contraction protocol (p > 0.22, all). Overall, the maintenance of PO2 above its critical threshold, as well as its dissociation from ATPOX, fatigue and the metabolites that cause it, indicate that an oxygen limitation does not impair bioenergetic pathways or power production during maximal knee extension contractions in vivo.