Purpose Different strategies for near-infrared spectroscopy (NIRS)-derived muscle oxidative capacity assessment have been reported. This study compared and evaluated (I) approaches for averaging trials; (II) NIRS signals and blood volume correction equations; (III) the assessment of vastus lateralis (VL) and tibialis anterior (TA) muscles in two fitness levels groups. Methods Thirty-six participants 18 chronically trained (CT: 14 males, 4 females) and 18 untrained (UT: 10 males, 8 females) participated in this study. Two trials of twenty transient arterial occlusions were performed for NIRS-derived muscle oxidative capacity assessment. Muscle oxygen consumption ( V ˙ O 2 m ) was estimated from deoxygenated hemoglobin (HHb), corrected for blood volume changes following Ryan (HHbR) and Beever (HHbB) equations, and from oxygen saturation (StO 2 ) in VL and TA. Results Superimposing or averaging V ˙ O 2 m or averaging the rate constants ( k ) from the two trials resulted in equivalent k values two one-sided tests (TOST) procedure with 5% equivalence margin— P  < 0.001. Whereas HHbR (2.35 ± 0.61 min −1 ) and HHbB (2.34 ± 0.58 min −1 ) derived k were equivalent ( P  < 0.001), StO 2 derived k (2.81 ± 0.92 min −1 ) was greater ( P  < 0.001) than both. k values were greater in CT vs UT in both muscles (VL: + 0.68 min −1 , P  = 0.002; TA: + 0.43 min −1 , P  = 0.01). Conclusion Different approaches for averaging trials lead to similar k . HHb and StO 2 signals provided different k , although different blood volume corrections did not impact k . Group differences in k were detected in both muscles.