This study examined the effects of blood-flow-restricted (BFR)-training on thigh glucose uptake at rest and during exercise in humans and the muscular mechanisms involved. Ten active men (~25 y; VO2max ~50 mL/kg/min) completed six weeks of training, where one leg trained with BFR (cuff pressure: ~180 mmHg) and the other leg without BFR. Before and after training, thigh glucose uptake was determined at rest and during exercise at 25% and 90% of leg incremental peak power output by sampling of femoral arterial and venous blood and measurement of femoral arterial blood flow. Furthermore, resting muscle samples were collected. After training, thigh glucose uptake during exercise was higher than before training only in the BFR-trained leg (p < 0.05) due to increased glucose extraction (p < 0.05). Further, BFR-training substantially improved time to exhaustion during exhaustive exercise (11 ± 5% vs. CON-leg; p = 0.001). After but not before training, NAC infusion attenuated (~50-100%) leg net glucose uptake and extraction during exercise only in the BFR-trained leg, which coincided with an increased muscle abundance of Cu/Zn-SOD (39%), GPX-1 (29%), GLUT4 (28%), and nNOS (18%) (p < 0.05). Training did not affect Mn-SOD, catalase, and VEGF abundance in either leg (p > 0.05), although Mn-SOD was higher in BFR-leg vs. CON-leg after training (p < 0.05). The ratios of p-AMPK-Thr172/AMPK and p-ACC-Ser79/ACC, and p-ACC-Ser79, remained unchanged in both legs (p > 0.05), despite a higher p-AMPK-Thr172 in BFR-leg after training (38%; p < 0.05). In conclusion, BFR-training enhances glucose uptake by exercising muscles in humans probably due to an increase in antioxidant function, GLUT4 abundance, and/or NO availability. Copyright © 2019 Elsevier Inc. All rights reserved.