Lung injury in ventilated premature infants occurs primarily through the mechanism of volutrauma, often due to the combination of high tidal volumes in association with a high end-inspiratory volume and occasionally end-expiratory alveolar collapse. Tolerating a higher level of arterial partial pressure of carbon dioxide (PaCO2) is considered as 'permissive hypercapnia' and when combined with the use of low tidal volumes may reduce volutrauma and lead to improved pulmonary outcomes. Permissive hypercapnia may also protect against hypocapnia-induced brain hypoperfusion and subsequent periventricular leukomalacia. However, extreme hypercapnia may be associated with an increased risk of intracranial hemorrhage. It may therefore be important to avoid large fluctuations in PaCO2 values. Recent randomized clinical trials in preterm infants have demonstrated that mild permissive hypercapnia is safe, but clinical benefits are modest. The optimal PaCO2 goal in clinical practice has not been determined, and the available evidence does not currently support a general recommendation for permissive hypercapnia in preterm infants.