Background: Skeletal muscle dysfunction and poor exercise tolerance are hallmarks of end-stage renal disease (ESRD). Noninvasively measured (near-infrared spectroscopy, NIRS) resting muscle oxygen consumption (rmVO2) is a biomarker of muscle dysfunction, which can be applied to study the severity and the reversibility of ESRD myopathy. We tested the hypothesis that deconditioning is a relevant factor in ESRD myopathy. Methods: The whole dialysis population (n = 59) of two of the eight centers participating into the EXCITE study (ClinicalTrials.gov NCT01255969), a randomized trial evaluating the effect of a home-based exercise program on the functional capacity of these patients was studied. Thirty-one patients were in the active arm (exercise group) and 28 in the control arm (no intervention). Normative data for rmVO2 were obtained from a group of 19 healthy subjects. Results: rmVO2 was twice higher (p < 0.001) in ESRDs patients (0.083 ± 0.034 ml/100 g/min) than in healthy subjects (0.041 ± 0.020 ml/100 g/min) indicating substantial skeletal muscle dysfunction in ESRD. rmVO2 correlated with resting heart rate (r = 0.34, p = 0.009) but was independent of age, dialysis vintage, biochemical, vascular and nutrition parameters. After the 6-month exercise program, rmVO2 reduced to 0.064 ± 0.024 ml/100 g/min (-23%, p < 0.001) in the exercise group indicating that skeletal muscle dysfunction is largely reversible but remained identical in the control group (0.082 ± 0.032 to 0.082 ± 0.031 ml/100 g/min). Conclusion: Deconditioning has a major role in ESRD myopathy. rmVO2 is a marker of physical deconditioning and has the potential for monitoring re-conditioning programs based on physical exercise in the ESRD population.