Selective reduction of pulmonary vascular resistance (PVR) remains a therapeutic goal for the treatment of pulmonary hypertension, but current therapeutic options remain limited. Although the gas nitric oxide (NO) selectively dilates the pulmonary vascular bed, it requires special equipment for administration, has a short biologic half-life, and is potentially toxic. We hypothesized that stimulation of the NO pathway at the level of its second messenger, guanosine 3',5'-cyclic monophosphate (cGMP), by targeted pulmonary delivery of a membrane-permeable nonhydrolyzable eGMP analogue would cause selective pulmonary vasodilation. Pulmonary hypertension was induced in 21 pigs by the intravenous infusion of a thromboxane A(2) analogue (9,11-dideoxy-9 alpha, 11 alpha-epoxymethanoprostaglandin F-2 alpha). Inhaled 8-bromoguanosine 3',5'-cyclic monophosphate (8-BrcGMP) lowered PVR in a time- and dose-dependent manner, with maximal effect achieved after 20 min, Compared with physiological saline control, 8-BrcGMP inhalation (3.0 mu g/kg) lowered PVR by 25 +/- 3% (P < 0.01), whereas there was no significant decline in systemic vascular resistance (4 +/- 6%); mean pulmonary arterial pressure declined 13 +/- 3% (P < 0.01), whereas there was little change in mean arterial pressure; cardiac output increased 10 +/- 4% (P < 0.05). PVR did not decrease after inhalation of noncyclic 8-bromoguanosine 5'-monophosphate, indicating that stimulation of the NO-cGMP pathway beyond the level of NO results in pulmonary vasodilation independent of stimulation of purinergic receptors. Inhaled 8-BrcGMP had no deleterious effect on load-independent measures of ventricular contractility, as shown by left ventricular pressure-volume loops generated at different preloads. Because selective pulmonary vasodilation was not observed after intravenous administration of 8-BrcGMP, these studies demonstrate that targeted delivery of a cGMP analogue by inhalation can selectively reduce PVR.