Treatments available for cerebral ischaemic stroke remain limited due to failures in clinical translation. To improve clinical translation, physiological oxygen encountered in vivo need to be considered in cell culture in vitro. Most in vitro studies are conducted under room air conditions (18 kPa O2); however, as cells in vivo experience O2 levels ranging from ~13 kPa to ~1 kPa, cells cultured under 18 kPa O2 are hyperoxic. Using an O2-sensitive probe (MitoXpress-INTRA, Agilent), we have established that intracellular O2 in mouse brain endothelial cells, bEnd.3, was 3.6 kPa during long-term culture under 5 kPa O2 in a Scitive O2-regulated workstation (Baker-Ruskinn), recapitulating reported intracellular O2 levels in the brain. We previously reported that long-term culture under 5 kPa O2, results in a phenotype different to cultures under 18 kPa O2, as evidenced by downregulation of specific Nrf2 target antioxidant genes. bEnd.3 cells (n=3-5) cultured either under 18 kPa O2 (hyperoxia) or 5 kPa O2 (physiological normoxia) were subjected to hypoxia (1 kPa O2, 1 h) and reoxygenation (back to either 18 or 5 kPa O2) to model in vivo ischaemic stroke. Reactive oxygen species (ROS) production was measured in real time in a CLARIOstar plate reader (BMG Labtech), using the chemiluminescent probe L-012 and mitochondria-specific ROS indicator, MitoSOX™ Red. Administration of a Nrf2 inducer, sulforaphane (2.5 µM), reduced the superoxide burst associated with reperfusion injury in cells cultured under 18 kPa O2. Furthermore, administration of rotenone (1 µM) 5 min prior to reoxygenation significantly reduced superoxide production compared to vehicle, implicating complex I of the electron transport chain in oxidative stress. Adaptation of bEnd.3 cells to 5 kPa O2 prevented superoxide production associated with ischaemia-reperfusion injury, suggesting that exaggerated response in cultures exposed to hyperoxia may potentially create misleading insights. Our findings highlight the importance of conducting vascular cell culture under physiological normoxia to recapitulate the redox phenotype of brain microvascular endothelial cells in vivo.