Blood flow generally decreases to areas of the lung that become hypoxie, thereby reducing pulmonary shunt and preventing arterial hypoxemia. The mechanism of blood flow reduction in the hypoxic area is believed to be hypoxic pulmonary vasoconstriction(HPV). Over the years, serious attention has been given to the question of whether hypoxia of the airway(alveolar hypoxia) is the only way to elicit pulmonary vasoconstriction. the current consensus favors this traditional view. Even though the usual experiments have succeeded in proving that a large drop in alveolar PO2(PAO2) does elicit pulmonary vasoconstriction, they have not excluded the possibility that a sufficient drop in mixed venous PO(PvO2) may have the same eonsequenee. The purpose of this study is to test the hypothesis that a decrease in PvO2 may elicit hypoxic pulmonary vasoconstriction. For the selective change of PvOwe used venovenous extracorporeal circulation(VVECMO), a technique which had been tested for adequacy in hemodynamics and oxygenation in preliminary studies on 6 mongrel dogs. In eight pentobarbital-anesthetized, mechanically ventilated mongrel dogs, stepwise reductions in oxygen concentration(F1O2,) from 0.21 to 0.15 and 0.10 caused remarkable HPV. The mean pulmonary arterial pressure(MPAP) and the pulmonary vascular resistance(PVR) were increased from 17.3±3 mmHg and 185±64 dyne sec cm(-5) in normoxia(F1O2=0.21) to 30±6 mmHg and 401±257 dyne sec cm in severe hypoxia(F1O2=0.10, P(A)O2=30±9 mmHg), respectively ; while, in moderate hypoxia(F1OP2=0.15, P(A)O2=53±12 mmHg), only the MPAP increased and the increase was from 17.3±3 mmHg to 23±3 mmHg. The selective increase in PvO2without any change in P(A)O2 by extracorporeal veno-venous bypass from 37±6 mmHg and 22±4 mmHg to 54±8 mmHg and 45±5 mnHg in moderate and severe hypoxia, respectively, depressed HPV: decrease in MPAP in moderate hypoxia(MPAP=21±4 mmHg) and decrease in MPAP and PVR in severe hypoxia(MPAP=23±5 mmHg, PVR=319±228 dyne sec * cm(-5)). This study indicates that the mixed venous oxygen tension is an important determinant of the magnitude of hypoxic pulmonary vasoconstriction in diffuse alveolar hypoxia. Moreover, the method used in this study and the result might be useful other pulmonary pathophysiological studies and clinical situations with acute diffuse alveolar hypoxia, such as acute respiratory failure and chronic obstructive pulmonary diseases.