This study evaluated the influence of cerebral blood flow on the decrease in frontal lobe oxygenation (ScO2) during maximal exercise by clamping end-tidal CO2 tension (PetCO2). Twelve endurance-trained males (age: 25±3 years, height: 183±8 cm, weight: 75±9 kg) performed, in a cross-over design, isocapnic (PetCO2 clamped at 40±1 mmHg) and poikilocapnic (ambient air) high-intensity constant load ergometer cycling to exhaustion. Mean blood flow velocity in the middle cerebral artery (MCA Vmean; transcranial Doppler ultrasound) and ScO2 (near-infrared spectroscopy) were assessed. Values are means±SD, compared by ANOVA. Exercise duration (12 min 24.9 s±1 min 18 s) was matched in the two trials, while pulmonary oxygen uptake (VO2) was not quite equal (90±6% vs. 93±5% of the maximal value; P=0.012) and PetCO2 was 40±1 and 34±4 mmHg, respectively (P<0.05). During the isocapnic trial MCA Vmean increased by 16±13% from the resting value until the time that clamping was applied and then increased further by 14±28% (P=0.017) until the end of exercise, while MCA Vmean did not change significantly during the poikilocapnic trial (P=0.071). In contrast, ScO2 decreased similarly (by -3.2±5.1% and -4.1±9.6%; P<0.001) at the end of both trials. During maximal exercise, ScO2 is reduced despite an increase in MCA Vmean suggesting dissociation between frontal lobe oxygenation and cerebral blood flow.