The relationship between skeletal muscle intracellular PO(2) (iPO(2)) and progressive muscular work has important implications for the understanding of O(2) transport and utilization. Presently there is debate as to whether iPO(2) falls progressively with increasing O(2) demand or reaches a plateau from moderate to maximal metabolic demand. Thus, using (1)H magnetic resonance spectroscopy of myoglobin (Mb), we studied cellular oxygenation during progressive single-leg knee extensor exercise from unweighted to 100% of maximal work rate in six active human subjects. In all subjects, the Mb peak at 73 ppm was not visible at rest, whereas the peak was small or indistinguishable from the noise in the majority of subjects during progressive exercise from unweighted to 50-60% of maximum work rate. In contrast, beyond this exercise intensity, a Mb peak of consistent magnitude was discernible in all subjects. When a Mb half saturation of 3.2 Torr was used, the calculated skeletal muscle PO(2) was variable before 60% of maximum work rate but in general was relatively high (>18 Torr, the measurable PO(2) with the poorest signal-to-noise ratio, in the majority of cases), whereas beyond this exercise intensity iPO(2) fell to a relatively uniform and invariant level of 3.8 +/- 0.5 Torr across all subjects. These results do not support the concept of a progressive linear fall in iPO(2) across increasing work rates. Instead, this study documents variable but relatively high iPO(2) from rest to moderate exercise and again confirms that from 50-60% of maximum work rate iPO(2) reaches a plateau that is then invariant with increasing work rate.