There has been considerable controversy surrounding the optimal inspired oxygen concentration for resuscitation of term and preterm infants. We have developed a rat pup model to quantify both physiologic and biochemical parameters associated with normoxic vs. hyperoxic resuscitation. We have confirmed existing human data that hyperoxic resuscitation of rat pups is associated with a significant delay in onset of spontaneous respiratory efforts. Both 40% and 100% inspired oxygen delayed onset of respiratory activity when compared to 21% oxygen. We have also documented, in the rat pup model, that hyperoxic resuscitation is associated with reduced levels of glutathione at 24 hours post resuscitation. The implications of these and other findings for human infants are that term asphyxiated babies can be safely resuscitated in 21% oxygen and that supplementary oxygen can be reserved for non-responders. In contrast, resuscitation of extremely low gestational age infants does appear to require an initial low inspired oxygen concentration (eg, 30%) with subsequent pulse oximetry titration to optimize oxygenation status.