Experimental studies to this point have not identified a selective neonatal pulmonary vasodilator. They have indicated that the neonatal pulmonary circulation is a complex, active vascular bed that has a number of endogenous vasodilatory mechanisms which oppose vasoconstriction under normal circumstances. It seems likely that a better understanding of how those mechanisms become deranged in various disease states will be required before we can substantially improve our drug therapy in pulmonary hypertensive infants. The data we have outlined above indicate that firm recommendations for drugs and their doses cannot be made. Nonetheless, several principles of therapy can be outlined. Because of the marginal benefits, which have resulted from current drug therapy [9, 24, 34, 79, 84, 100, 107], it seems clear that, at the moment, the most prudent initial course in neonates with pulmonary vasospasm should be nonpharmacologic: restoration of normal blood gases, use of high concentrations of inspired oxygen with hyperventilation to pH 7.6 if cyanosis persists [22, 79], avoidance of agitation and hypothermia [22], and correction of any metabolic derangements [92]. Decreased cardiac output should be identified and treated with blood volume expanders and/or cardiotonic agents as necessary. Finally, if physiologic efforts to lower pulmonary vascular resistance fail, drug therapy sometimes is helpful in effecting salvage. Treated infants should be carefully monitored, not only for signs of improved oxygenation, but also for changes in right-to-left ductal shunting and cardiac output. If a given agent does not produce beneficial effects at a range of doses by 60 min, it is unlikely that prolonged therapy will result in late improvement. In such circumstances, a change in drug therapy is probably indicated. Finally, it would seem wise to use multiple agents with extreme cautions, being careful to pair a direct vasodilator (e.g., nitroprusside, tolazoline, or prostacyclin) with a cardiotonic agent (e.g., isoproterenol or dopamine), or simultaneously administered volume expanders as the individual clinical situation dictates. Further animal experimentation will undoubtedly identify new and promising agents and provide an increasing understanding of the cellular physiology of the newborn pulmonary circulation. However, only careful clinical and experimental studies into the cause(s) of the vasoconstriction in newborns will allow the development of a truly rational approach to specific therapy in the human species.