Objective: The aim of our study was to examine cognition response to sprint interval exercise (SIE) against different levels of hypoxia. Research design and methods: 26 recreational active males performed SIE (20 × 6 s of all-out cycling bouts, 15 s of passive recovery) under normoxia (FO: 0.209), moderate hypoxia (FO: 0.154), and severe hypoxia (FO: 0.112) in a single-blinded crossover design. Cognitive function and blood glucose were assessed before and after 0, 10, 30, and 60 min of the SIE. Heart rate (HR), peripheral oxygen saturation (SpO), and ratings of perceived exertion (RPE, the Borg 6–20-point scale) during each SIE trial were recorded before and immediately after every five cycling bouts, and after 0, 10, 30, and 60 min of the SIE. Results: All the three SIE trials had a significantly faster overall reaction time in the Stroop test at 10 min after exercise as compared to that of the baseline value (p = 0.003, η = 0.606), and returned to normal after 60 min. The congruent RT at 10 min after SIE was significantly shorter than that of the baseline (p < 0.05, η = 0.633), while the incongruent RT at both 10 min and 30 min were significantly shorter than that measured at baseline (p < 0.05, η = 0.633). No significant differences in terms of accuracy were found across the three trials at any time points (p = 0.446, η = 0.415). Blood glucose was significantly reduced at 10 min and was sustained for at least 60 min after SIE when compared to pre-exercise in all trials (p < 0.05). Conclusions: Acute SIE improved cognitive function regardless of oxygen conditions, and the sustained improvement following SIE could last for at least 10–30 min and was unaffected by the altered blood glucose level.