Background. Severe bronchopulmonary dysplasia (BPD), which is associated with high mortality and morbidity, is thought to be the result of mechanical, inflammatory, and oxidant injury to the immature lung, and includes the development of pulmonary hypertension with vascular remodeling. Methods. A phase II pilot study was conducted to determine the effect of inhaled nitric oxide (iNO) on oxygenation in severe BPD. This was an open-labeled, noncontrolled trial to evaluate safety and determine appropriate dosing for a future randomized controlled trial. Infants were eligible for enrollment if they were greater than or equal to 4 weeks of age and ventilator dependent with a mean airway pressure of greater than or equal to 10 cm H2O and an FIO2 of greater than or equal to 0.45. Study infants received iNO (20 ppm) for 72 hours, and FIO2 was adjusted to maintain oxygen saturations of >92%. Infants who had a greater than or equal to 15% reduction in FIO2 after 72 hours received prolonged treatment with low-dose iNO, weaning by 20% every 3 days as tolerated. Findings. Sixteen preterm infants (23-29 weeks of gestation), age 1 to 7 months, were enrolled. Eleven of 16 infants had a significant increase in Fao, after 1 hour of iNO (median change, 24 mm Hg; range, -15 to 59 mm Hg; P <.01), but there was no significant change in PaCO2. After 72 hours of iNO, 11 infants had greater than or equal to 15% reduction in FIO2, and 7 of the 11 had greater than or equal to 35% reduction (P <.01). Among the 11 infants who responded to iNO after 72 hours, 10 had a sustained improvement in oxygenation throughout their course of treatment (duration, 8-90 days), and ventilator support could also be decreased. No adverse effects from iNO (increased methemoglobin, bleeding, or increased plasma 3-nitrotyrosine) were observed. Four of the 11 infants (36%) who responded to iNO ultimately weaned off mechanical ventilation and 4 died, whereas all the infants who failed to respond to iNO either died or continue to require mechanical ventilation. Interpretation. We conclude that the use of low-dose iNO may improve oxygenation in some infants with severe BPD, allowing decreased FIO2 and ventilator support without evidence of adverse effects. Randomized clinical trials of low-dose iNO for BPD are warranted.