There is evidence that the concentration of the high-energy phosphate metabolites may be altered during steady-state submaximal exercise by the breathing of different fractions of inspired O-2 (FIO2). Whereas it has been suggested that these changes may be the result of differences in time taken to achieve steady-state O-2 uptake (Vo(2)) at different FIO2 values, we postulated that they are due to a direct effect of O-2 tension. We used P-31-magnetic resonance spectroscopy during constant-load, steady-state submaximal exercise to determine 1) whether changes in high-energy phosphates do occur at the same Vo(2) with varied FIO2 and 2) that these changes are not due to differences in Vo(2) onset kinetics. Six male subjects performed steady-state submaximal plantar flexion exercise [7.2 +/- 0.6 (SE) W] for 10 min while lying supine in a 1.5-T clinical scanner. Magnetic resonance spectroscopy data were collected continuously for 2 min before exercise, 10 min during exercise, and 6 min during recovery. Subjects performed three different exercise bouts at constant load with the FIO2 switched after 5 min of the 10-min exercise bout. The three exercise treatments were 1) FIO2 of 0.1 switched to 0.21, 2) FIO2 of 0.1 switched to 1.00, and 3) FIO2 of 1.00 switched to 0.1. For all three treatments, the Fro, switch significantly (P less than or equal to 0.05) altered phosphocreatine: 1) 55.5 +/- 4.8 to 67.8 +/- 4.9% (%rest); 2) 59.0 +/- 4.3 to 72.3 +/- 5.1%; and 3) 72.6 +/- 3.1 to 64.2 +/- 3.4%, respectively. There were no significant differences in intracellular pH for the three treatments. The results demonstrate that the differences in phosphocreatine concentration with varied FIO2 are not the result of different Vo(2) onset kinetics, as this was eliminated by the experimental design. These data also demonstrate that changes in intracellular oxygenation, at the same work intensity, result in significant changes in cell homeostasis and thereby suggest a role for metabolic control by O-2 even during submaximal exercise.