We investigated whether resuscitation with lower O2 concentrations than 21% reduces reoxygenation injury and thereby improves cerebral metabolic recovery and EEG recovery. Methods: Hypoxia was induced by ventilating 24 anesthetized and instrumented newborn piglets with 6% O2 in N2. CO2 was added to the inspiratory gas to achieve PaCO2 between 52 and 60 mm Hg. When EEG became isoelectric, and either mean arterial blood pressure fell below 25 mmHg or base excess (BE) was lower than -25 mmol/L, a 2 hour resuscitation period was started. The piglets were randomly divided into three groups: Hypoxemic (n=8), 21% O2 (n=8), and 100% O2 group (n=8). The hypoxemic group was ventilated with 12 to 18% O2 to achieve a cerebral venous O2 saturation of 17 to 23%(baseline; 42-46%), levels which in a previous study was shown not to give any changes of EEG or hypoxanthine (Hx) in cerebral cortex in healthy piglets. PaCO2 was kept in normal range during resuscitation. Extracellular Hx was sampled by in vivo microdialysis from cerebral cortex and striatum. EEG and Oxygenation Index (Near Infrared Spectrophotometry, NIRS) were continuously recorded. Results: Arterial oxygen tensions at 30 mins of resuscitation were (mean±SEM) 45±2, 74±4, and 431±20 mmHg in the hypoxemic, 21% O2 and 100% O2 groups. There were no significant differences between the groups in recovery of EEG(p=0.22) or Hx concentrations in cerebral cortex, striatum, and plasma. Mean BE was -20 to -23 mmol/L at the end of hypoxia (p=0.67) and rose during resuscitation to -13±2, -5±2, and -5±1 mmol/L in the hypoxemic, 21% O2 and 100% O2 groups(p<0.05, hypoxemic vs. 21% O2 and 100% O2 groups). Resuscitation with 21% O2 and 100% O2 reestablished the O2 support to the brain measured by NIRS within the first mins of resuscitation. Conclusion: Hypoxemic resuscitation of newborn piglets does not improve cerebral metabolic recovery or EEG recovery, and delays the correction of the metabolic acidosis.