We investigated how blood flow restriction (BFR) training enhances performance and neuromuscular function during dynamic exercise through adaptations in vascular function, mitochondrial respiratory capacity and metaboreflex-related haemodynamic responses. Eleven healthy active participants completed 6 weeks of interval training with one leg under BFR (BFR-leg; cuff pressure:165 ± 36 mmHg) and the other without (CTRL-leg). Before and after training, participants performed single-leg knee-extensor incremental tests to determine peak power output (Wmax) and constant workload trials at 90% pre-training Wmax to assess neuromuscular fatigue [maximal voluntary contraction (MVC) and evoked contractile force (QT)]. Blood flow responsiveness was evaluated using Doppler ultrasound during reactive hyperaemia (RH), passive leg movement (PLM) and rapid-onset vasodilatation (ROV). Mitochondrial respiratory capacity was estimated indirectly from post-exercise recovery kinetics of muscle oxygen consumption (mdot V_mathrmO_2). Mean arterial pressure (MAP) and heart rate (HR) responses to metaboreflex activation during exercise and post-exercise circulatory occlusion (PECO) were measured. After training, Wmax increased ∼17% more in the BFR-leg than in the CTRL-leg (P textless 0.001). Exercise-induced reductions in MVC (P textless 0.001) and QT forces (P = 0.001) were less in the BFR-trained leg than CTRL-leg, indicating less fatigue development. Femoral blood flow was higher in the BFR-leg during RH and ROV (P textless 0.001), but not PLM. MAP (P = 0.001) and HR (P textless 0.001) responses were lower in the BFR-leg during exercise with muscle metaboreflex activation, without training or leg differences during PECO (P textgreater 0.230). The kinetics of mdot V_mathrmO_2 improved similarly in both legs (P textless 0.007). Compared to the control, BFR interval training elicited greater improvements in performance and less fatigue for equivalent exercise through integrated neuromuscular, vascular and metabolic adaptations. Key points Interval training with blood flow restriction (BFR) improves exercise performance by enhancing oxygen delivery and muscle homeostasis in contracting muscles, but its effects on neuromuscular fatigue and the underlying mechanisms are not fully understood. In a within-subject, unilateral design, participants trained one leg with BFR for 6 weeks while the contralateral leg served as a control. Before and after training, we assessed neuromuscular fatigue, vascular function, near-infrared spectroscopy (NIRS)-derived post-exercise recovery kinetics of muscle oxygen consumption (mdot V_mathrmO_2, a surrogate for mitochondrial respiratory capacity) and muscle metaboreflex activation. BFR interval training attenuated neuromuscular fatigue development during high-intensity dynamic knee-extensor exercise, mainly by reducing impairments in peripheral (i.e. muscle) fatigue. After BFR training, leg blood flow was greater during vascular function tests compared to the control, suggesting enhanced haemodynamic function. Less fatigue for equivalent exercise occurred independently of changes in NIRS-derived mdot V_mathrmO_2 and appears to be largely driven by peripheral neuromuscular and vascular adaptations.