Introduction Continuous wave (CW) near-infrared spectroscopy (NIRS) is widely used in skeletal muscle tissue hemodynamic and respiratory research utilizing fixed light scattering assumptions. Purpose Compare reproducibility and technical characteristics of CW-NIRS against frequency-domain (FD)-NIRS, which allows for intra-individual scattering correction. Methods Muscle blood flow (mBF) and oxygen uptake ( ) (using rapid-occlusions), and perfusion were assessed with NIRS in 10 healthy (ND) and 10 type-2 diabetic (T2D) men at rest and during knee extensions at 5% and 15% of maximal voluntary contraction (MVC) on three occasions within 10 days. Results In both groups, exercise responses for mBF and were not substantially different between technology with moderate-high between-day reproducibility (ICC range 0.72–0.98). However, compared to FD-NIRS, CW-NIRS perfusion substantially higher at rest (21 uM; 90%CI 29–13), 5%MVC (18 uM; 27–9) and 15%MVC (17 uM; 26–8) in T2D, but not ND. Standardized typical errors during exercise were small-moderate: e.g., 15%MVC mBF, CW-NIRS (ND 0.45; T2D 0.49), FD-NIRS (0.35; 0.49); mV̇O2, CW-NIRS (0.24; 0.26), FD-NIRS (0.45; 0.32); perfusion CW-NIRS (0.25; 0.25) and FD-NIRS (0.25; 0.23); and, not different between groups and technology. Conclusion Both NIRS technologies are reproducible. CW-NIRS assumptions work for muscle blood flow and oxygen use at rest and during exercise. FD-NIRS might be better at detecting changes in perfusion during exercise, especially in older or clinical groups. However, FD-NIRS is more expensive and complex, making it less practical.