Dexamethasone (DEX) pretreatment has been shown to be neuroprotective in a neonatal rat model of hypoxia ischemia (HI). The exact mechanism of this neuroprotection is still unknown. This study used 31P nuclear magnetic resonance spectroscopy to monitor energy metabolism during a 3-h episode of HI in 7-d-old rat pups in one of two groups. The first group was pretreated with 0.1 mL saline (i.p.) and the second group was treated with 0.1 mL of 0.1mg/kg DEX (i.p.) 22 h before HI. Animals pretreated with DEX had elevated nucleoside triphosphate and phosphocreatine levels during HI when compared with controls. Saline-treated animals had significant decreases in nucleoside triphosphate and phosphocreatine and increases in inorganic phosphate over this same period. 31P nuclear magnetic resonance data unequivocally demonstrate preservation of energy metabolism during HI in neonatal rats pretreated with DEX. Animals pretreated with DEX had little or no brain damage following 3 h of HI when compared with matched controls, which experienced severe neuronal loss and cortical infarction. These same pretreated animals had an increase in blood beta-hydroxybutyrate levels before ischemia, suggesting an increase in ketone bodies, which is the neonate's primary energy source. Elevation of ketone bodies appears to be one of the mechanisms by which DEX pretreatment provides neuroprotection during HI in the neonatal rat.