To test our hypothesis that resuscitation with 21% and 40% oxygen (O2) would shorten time to onset of respiratory activity when compared with resuscitation with 100% O2, diaphragmatic electromyogram (EMG) electrodes were inserted in Sprague-Dawley rat pups, age 8-10 d before intubation and mechanical ventilation with 5% O2 to induce cessation of respiratory activity. Each animal was then resuscitated with 100% and 21% O2 (n = 10) or 100% and 40% O2 (n = 11) for 30 s before the ventilator was disconnected. Recovery of diaphragm activity was compared between resuscitation groups. Blood gas status and heart rate data were characterized in additional rat pups. Time to first respiratory effort was 36 +/- 21 s (mean +/- SD) for room air resuscitation and 72 +/- 22 s for 100% O2, (p = 0.002). In contrast, there was no difference in time to onset of diaphragm activity when resuscitation with 40% O2 was compared with 100% O2: 84 +/- 27 s versus 76 +/- 23 s, respectively (p > 0.05). Resuscitation with 100% and 40% O2 both resulted in hyperoxia and hypocapnia when compared with room air, without effect on heart rate. Our findings indicate that even modest hyperoxic resuscitation will result in a delayed onset of respiration compared with normoxic gas, via a mechanism that may involve both hyperoxemic and hypocapnic inhibition of chemoreceptors.