Purpose <p>Acute hypoxia can impair processing speed and visual search. In this exploratory non-randomised controlled study, we examined whether normoxic brain endurance training (BET) is associated with cognitive and oculomotor performance during acute hypoxic exposure.</p> Methods <p>Thirteen adults (7 females) completed a 6-week BET programme, while 23 separately recruited control participants (10 females) did not. Both groups were tested under normobaric hypoxia (fraction of inspired oxygen, FiO₂ = 0.14, simulated 3,200 m). Participants performed a psychomotor vigilance task (PVT), visual search task (VST) with eye tracking, and stop-signal task (SST) before and after a 20-min Stroop task. Linear mixed-effects models (LMMs) tested Group × Time effects; propensity-score sensitivity analyses were conducted.</p> Results <p>Compared with controls, BET was associated with faster PVT reaction time (<i>p</i> &lt; 0.05, d = − 0.84) and faster VST target and non-target reaction times (both <i>p</i> &lt; 0.001; d = − 2.45 and − 1.70), while maintaining accuracy; VST target accuracy was higher (<i>p</i> &lt; 0.05, d = 0.89). SST outcomes showed no group differences (all <i>p</i> &gt; 0.05). Pupil diameter increased from baseline to post-fatigue (<i>p</i> &lt; 0.05, d = 0.81) with no clear group difference. Saccade duration showed a Group×Time interaction (<i>p</i> &lt; 0.001, partial η² = 0.29), with a larger post-fatigue reduction in the BET group; sensitivity analyses showed consistent directions.</p> Conclusion <p>Normoxic BET was associated with faster processing speed and altered oculomotor behaviour under acute hypoxia, while inhibitory control was unaffected. Findings are exploratory and should be interpreted cautiously given the non-randomised and imbalanced design.</p> Clinical trial registration <p>This study was not prospectively registered.</p> Graphical abstract <p></p>