Significance: Near‐infrared spectroscopy (NIRS) measurement is a widely used technique to measure muscle oxygenation. A knowledge of the reproducibility of NIRS measurements is essential for the correct interpretation of data. Aim: Our aim was to test the reproducibility and sex differences of NIRS measurements during brachial artery occlusion in healthy participants. Approach: An NIRS device was used to measure muscle oxygenation and microvascular function during a 5 min brachial occlusion. Muscle oxygen consumption (mVO2) and tissue saturation index (TSI%) were used. The occlusion test was performed three times on separate days for males (n = 13, 28 ± 8 years) and females (n = 13, 29 ± 7 years). Results: During the occlusion phase, the reproducibility of mVO2 was excellent (intraclass correlation; ICC = 0.90). During the reperfusion phase, the maximal change in TSI% revealed the best reproducibility (ICC = 0.77). There were no sex differences in reproducibility. Male participants had higher muscle oxygenation during occlusion (mVO2, 0.054 ± 0.010 vs. 0.038 ± 0.012 mLO2/min/100 g, p = 0.001, male and female, respectively). There were no sex differences during the reperfusion phase. Conclusion: The reproducibility of NIRS to measure muscle oxygenation and microvascular function during circulation occlusion and reperfusion is good to excellent. Muscle oxygen capacity measured during occlusion is higher in males compared to females, and there are no sex differences in microvascular function during the reperfusion phase.