The hypothesis that endothelium-derived relaxing factor (EDRF) modulates hypoxic pulmonary vasoconstriction (HPV) was tested in isolated, blood-perfused rat lungs ventilated with gas mixtures of 21% O2-5% CO2-74% N2 (normoxia) or of 3% O2-5% CO2-92% N2 (hypoxia); 30-mu-M N(G)-monomethyl-L-arginine (L-NMMA), an Inhibitor of EDRF production, caused a reduction in the endothelium-dependent relaxant response to acetylcholine (ACh) from 62 +/- 7, 88 +/- 4, and 100 +/- 4% to 26 +/- 8, 49 +/- 12, and 75 +/- 7% at ACh concentrations of 1, 10, and 100-mu-M, respectively (p < 0.05 at all concentrations), indicating that L-NMMA acts via the inhibition of EDRF production. L-NMMA induced a concentration-related augmentation in HPV of 20 +/- 5, 32 +/- 8, and 34 +/- 8% at concentrations of 30, 300, and 1,000-mu-M (p < 0.05, compared with a vehicle control group at all concentrations). The pressor response to a dose of angiotensin II (A-II), which produced the same increase in pulmonary artery pressure as that induced by hypoxia, was also significantly augmented (2 +/- 0.6%), but to a lesser extent. The augmentation of HPV by 30-mu-M L-NMMA was completely reversed by 1 mM L-arginine (a precursor of EDRF), but not by D-arginine (an isomer of L-arginine). One of 6 mM L-arginine, but not 6 mM D-arginine caused a significant inhibition of HPV by 20 +/- 2 and 47 +/- 12% (p < 0.05, compared with the vehicle control group) and a small but not significant reduction in A-II-mediated contraction. We conclude that endothelium-derived relaxing factor modulates HPV and that the modulating action of EDRF on pulmonary blood pressure may be more relevant to HPV than to A-II-induced concentration. The ability of L-arginine to reduce HPV suggests that the formation of EDRF during hypoxemia may be limited by the availability of its precursor.