We investigated the effects of nitroprusside and isoflurane on multipoint pulmonary arterial pressure (PAP)/cardiac index (Q) plots in pentobarbital sodium-anesthetized dogs ventilated alternatively in hyperoxia (fraction of inspired O2, FIO2, 0.4) and hypoxia (FIO2 0.1). Over the entire range of Q studied, 2-5 l.min-1.m-2, hypoxia increased PAP in 16 dogs ("responders") and did not affect PAP in 16 other dogs ("nonresponders"). A hypoxic pulmonary vasoconstriction (HPV) was restored in the nonresponders by intravenous administration of 1 g of acetylsalicylic acid (ASA). Nitroprusside (5 micrograms.kg-1.min-1) inhibited HPV in responders (n = 8) and nonresponders treated with ASA (n = 8). End-tidal 1.41% isoflurane (a minimal alveolar concentration equal to one for dogs) did not affect HPV in responders (n = 8) and nonresponders treated with ASA (n = 8). In the latter group isoflurane increased PAP at the highest Q studied (3-5 l.min-1.m-2) in hyperoxia and hypoxia. In a final group of eight dogs with Q kept constant, PAP remained unchanged during two consecutive sequences of alternated 30-min periods (maximum time to generate a PAP/Q plot) successively at FIO2 0.4 and 0.1, and the hypoxia-induced increase in PAP was reproducible. Thus the present experimental model appeared suitable for the study of the effects of hypoxia and drugs on pulmonary vascular tone of intact dogs. At the given doses HPV was inhibited by nitroprusside and not affected by isoflurane. Products of arachidonic acid metabolism possibly could be implicated in the pulmonary vascular effects of isoflurane.