BACKGROUND: Prostacyclin administered intravenously has demonstrated intermediate pulmonary specificity and its aerosol form has an even greater pulmonary selectivity. There have been few systematic analyses of the difference in response according to the route of administration and the dose of administration of prostacyclin. So we have compared prostacyclin infusion versus inhalation in various concentrations in an animal model. METHODS: Pulmonary hypertension was induced by continuous intravenous infusion of the vasoconstrictor U46619 and prostacyclin solutions of 10, 50, 100, 200 mcg/ml were inhaled using a jet nebulizer. Prostacyclin infusion was done at a rate of 100, 200, 400 ng/kg/min. RESULTS: With inhalation of 10, 50, 100, 200 mcg/ml prostacyclin, PVR fell to values of 85%, 76%, 64%, 55% of the preinhalation value and SVR fell to values of 94%, 80%, 76%, 64% of the preinhalation value, respectively (p<0.05). PVR/SVR ratios decreased significantly in all inhalation doses (p<0.05). With infusion of prostacyclin at a rate of 100, 200, 400 ng/kg/min, PVR fell to values of 73%, 60%, 50% of the preinfusion value and SVR fell to values of 68%, 54%, 38% of the preinfusion value, respectively (p<0.05). PVR/SVR ratios increased at an infusion rate of 400 ng/kg/min. CONCLUSION: Prostacyclin inhalation did not result in selective pulmonary vasodilation without causing any efects on the systemic vascular bed (absolute pulmonary selectivity). But it did cause more predominant vasodilation on the pulmonary vascular bed (relative pulmonary selectivity). By contrast, prostacyclin infusion caused more predominant vasodilation on the systemic vascular bed, creating the risk of severe systemic hypotension.