Inhaled nitric oxide (iNO) has been shown to improve oxygenation in severe persistent pulmonary hypertension of the newborn (PPHN). However, PPHN is often associated with various lung diseases. Thus, response to iNO may depend upon the aetiology of neonatal acute respiratory failure. A total of 150 (29 preterm and 121 term) newborns with PPHN were prospectively enrolled on the basis of oxygenation index (OI) higher than 30 and 40, respectively. NO dosage was stepwise increased (10-80 ppm) during conventional mechanical or high-frequency oscillatory ventilation while monitoring the oxygenation. Effective dosages ranged from 5 to 20 ppm in the responders, whereas iNO levels were unsuccessfully increased up to 80 ppm in the nonresponders. Within 30 min of iNO therapy, OI was significantly reduced in either preterm neonates (51+/-21 vs 23+/-17, P < .0001) or term infants with idiopathic or acute respiratory distress syndrome (45+/-20 vs 20+/-17, P < .0001), 'idiopathic' PPHN (39+/-14 vs 14+/-9, P < .0001), and sepsis (55+/-25 vs 26+/-20, P < .0001) provided there was no associated refractory shock. Improvement in oxygenation was less significant and sustained (OI=41+/-16 vs 28+/-18, P < .001) in term neonates with meconium aspiration syndrome and much less (OI=58+/-25 vs 46+/-32, P < .01) in those with congenital diaphragmatic hernia. Only 21 of the 129 term newborns (16%) required extracorporeal membrane oxygenation (57% survival). Survival was significantly associated with the magnitude in the reduction in OI at 30 min of iNO therapy, a gestational age > or =34 weeks, and associated diagnosis other than congenital diaphragmatic hernia. Conclusion, iNO improves the oxygenation in most newborns with severe hypoxaemic respiratory failure including preterm neonates. However, response to iNO is disease-specific. Furthermore, iNO when combined with adequate alveolar recruitment and limited barotrauma using exogenous surfactant and HFOV may obviate the need for extracorporeal membrane oxygenation in many term infants.