Key points: It is unknown whether excessive reactive oxygen species (ROS) production drives the isocapnic hyperoxia (IH)-induced decline in human cerebral blood flow (CBF) via reduced nitric oxide (NO) bioavailability and leads to disruption of the blood–brain barrier (BBB) or neural-parenchymal damage. Cerebral metabolic rate for oxygen (CMR _O2 ) and transcerebral exchanges of NO end-products, oxidants, antioxidants and neural-parenchymal damage markers were simultaneously quantified under IH with intravenous saline and ascorbic acid infusion. CBF and CMR _O2 were reduced during IH, responses that were followed by increased oxidative stress and reduced NO bioavailability when saline was infused. No indication of neural-parenchymal damage or disruption of the BBB was observed during IH. Antioxidant defences were increased during ascorbic acid infusion, while CBF, CMR _O2 , oxidant and NO bioavailability markers remained unchanged. ROS play a role in the regulation of CBF and metabolism during IH without evidence of BBB disruption or neural-parenchymal damage. Abstract: To test the hypothesis that isocapnic hyperoxia (IH) affects cerebral blood flow (CBF) and metabolism through exaggerated reactive oxygen species (ROS) production, reduced nitric oxide (NO) bioavailability, disturbances in the blood–brain barrier (BBB) and neural-parenchymal homeostasis, 10 men (24 ± 1 years) were exposed to a 10 min IH trial (100% O ₂ ) while receiving intravenous saline and ascorbic acid (AA, 3 g) infusion. Internal carotid artery blood flow (ICABF), vertebral artery blood flow (VABF) and total CBF (tCBF, Doppler ultrasound) were determined. Arterial and right internal jugular venous blood was sampled to quantify the cerebral metabolic rate of oxygen (CMR _O2 ), transcerebral exchanges (TCE) of NO end-products (plasma nitrite), antioxidants (AA and AA plus dehydroascorbic acid (AA+DA)) and oxidant biomarkers (thiobarbituric acid-reactive substances (TBARS) and 8-isoprostane), and an index of BBB disruption and neuronal-parenchymal damage (neuron-specific enolase; NSE). IH reduced ICABF, tCBF and CMR _O2 , while VABF remained unchanged. Arterial 8-isoprostane and nitrite TCE increased, indicating that CBF decline was related to ROS production and reduced NO bioavailability. AA, AA+DA and NSE TCE did not change during IH. AA infusion did not change the resting haemodynamic and metabolic parameters but raised antioxidant defences, as indicated by increased AA/AA+DA concentrations. Negative AA+DA TCE, unchanged nitrite, reductions in arterial and venous 8-isoprostane, and TBARS TCE indicated that AA infusion effectively inhibited ROS production and preserved NO bioavailability. Similarly, AA infusion prevented IH-induced decline in regional and total CBF and re-established CMR _O2 . These findings indicate that ROS play a role in CBF regulation and metabolism during IH without evidence of BBB disruption or neural-parenchymal damage.