Since the discovery of retrolental fibroplasia, and the role of oxygen in its development, oxygen has been considered a double-edged sword in neonatal medicine, the utmost care being exercised in order not to give too much oxygen (1). However, the important observation that hypoxaemia might induce pulmonary vasoconstriction (2) and airway constriction (3) in infants at risk for bronchopulmonary dysplasia has resulted in only a minor upward adjustment of oxygen supplementation in many neonatal units. Since oxygen toxicity has long been linked not only to retinopathy of prematurity but also to bronchopulmonary dysplasia (4), it is relevant to ask whether an increased FiO2 might have any detrimental effects on babies.