The study examined the influence of cerebral (prefrontal cortex) and muscle (vastus lateralis) oxygenation on the ability to perform repeated, cycling sprints. Thirteen team-sport athletes performed ten, 10-s sprints (with 30 s of rest) under normoxic (F(I)O(2) 0.21) and acute hypoxic (F(I)O(2) 0.13) conditions in a randomised, single-blind fashion and crossover design. Mechanical work was calculated and arterial O(2) saturation (S(p)O(2)) was estimated via pulse oximetry for every sprint. Cerebral and muscle oxy-(O(2)Hb), deoxy-(HHb), and total haemoglobin (THb) were monitored continuously by near-infrared spectroscopy. Compared with normoxia, hypoxia induced larger decrements in S(p)O(2) and work (11.6 and 7.6%, respectively; P < 0.05). In the muscle, we observed a fairly constant level of deoxygenation across sprints, with no effect of the condition. In normoxia, regional cerebral oxygenation increased during the first two sprints and slightly fluctuated thereafter. In contrast, this initial cerebral hyper-oxygenation was attenuated in hypoxia. Changes in [O(2)Hb] and [HHb] occurred earlier and were larger in hypoxia compared with normoxia (P < 0.05), while regional blood volume (Delta[THb]) remained unaffected by the condition. Changes in cerebral [HHb] and mechanical work were strongly correlated in normoxia and hypoxia (R(2) = 0.81 and R(2) = 0.85, respectively; P < 0.05), although the slope of this relationship differed (normoxia, -351.3 +/- 183.3 vs. hypoxia, -442.4 +/- 227.2; P < 0.05). The results of this NIRS study show that O(2) availability influences prefrontal cortex, but not muscle, oxygenation during repeated, short sprints. By using a hypoxia paradigm, the study suggests that cerebral oxygenation contributes to the impairment of repeated-sprint ability.