The authors studied the influence of locally administered isoflurane anesthesia on the pulmonary vascular response to regional alveolar hypoxia (hypoxic pulmonary vasoconstriction [HPV]) over a range of cardiac outputs (COs) in seven mechanically ventilated, closed-chest dogs. The right lung was ventilated with 100% O2 throughout the study. The left lung was ventilated with either 100% O2 (normoxia) or an hypoxic gas mixture (hypoxia). Different alveolar concentrations of isoflurane (0, 1, and 2.5 MAC) were administered to the left lung in a randomized sequence. The CO was altered by opening and closing surgically produced arteriovenous fistulae, at all isoflurane concentrations, and by hemorrhage at 0 MAC isoflurane. The magnitude of the HPV response was measured by differential CO2 elimination in the absence of isoflurane and by venous admixtures in all phases. During normoxia, the left lung effective flow (QL%) measured from differential CO2 excretion was 39.9 +/- 1.2% of the total blood flow and decreased to 18.8 +/- 2.6% when ventilated with the hypoxic gas mixture. Venous admixture (QVA/QT%) was significantly correlated with QL% during hypoxic ventilation in the absence of isoflurane. QVA/QT% was 22.3 +/- 2.7% during hypoxia with normal CO, and it increased significantly to 27.7 +/- 1.1% when the CO was increased 43%. It was not significantly altered (23.6 +/- 3.6%) when the CO was decreased by 54%. Isoflurane 2.5 MAC significantly increased QVA/QT% during hypoxic ventilation of the left lung to 33.9 +/- 2.6% with low CO and 35.4 +/- 1.7% with normal CO. Isoflurane 1 MAC increased QVA/QT% to 27.2 +/- 2.7% with normal CO and 28.1 +/- 2.6% with high CO.(ABSTRACT TRUNCATED AT 250 WORDS)