This study compared the kinetics of muscle deoxygenation and reoxygenation during a sprint interval protocol performed under four modalities: blood flow restriction at 60% of the resting femoral artery occlusive pressure (BFR), gravity-induced BFR (G-BFR), simulated hypoxia (FiO ₂ ≈13%, HYP) and normoxia (NOR). Thirteen healthy men performed each session composed of five all-out 30-s efforts interspaced with 4 min of passive recovery. Total work during the exercises was 17 ± 3.4, 15.8 ± 2.9, 16.7 ± 3.4, and 18.0 ± 3.0 kJ for BFR, G-BFR, HYP and NOR, respectively. Muscle oxygenation was continuously measured with near-infrared spectroscopy. Tissue saturation index (TSI) was modelled with a linear function at the beginning of the sprint and reoxygenation during recovery with an exponential function. Results showed that both models were adjusted to the TSI (R ² = 0.98 and 0.95, respectively). Greater deoxygenation rates were observed in NOR compared to BFR (p = 0.028). No difference was found between the conditions for the deoxygenation rates relative to sprint total work (p > 0.05). Concerning reoxygenation, the amplitude of the exponential was not different among conditions (p > 0.05). The time delay of reoxygenation was longer in BFR compared to the other conditions (p < 0.05). A longer time constant was found for G-BFR compared to the other conditions (p < 0.05), and mean response time was longer for BFR and G-BFR. Finally, sprint performance was correlated with faster reoxygenation. Hence, deoxygenation rates were not different between the conditions when expressed relatively to total sprint work. Furthermore, BFR conditions impair reoxygenation: BFR delays and G-BFR slows down reoxygenation.