We examined the mechanism(s) of hypotensive action of platelet-activating factor (PAF) in anesthetized dogs. PAF (0.5 micrograms/kg i.v.) caused a biphasic hypotension; the first phase was transient and was accompanied by a decrease in systemic vascular resistance and an increase in cardiac output. Aspirin-DL-lysine, a cyclooxygenase inhibitor, had no effect on this phase. The second phase was characterized by a sustained hypotension caused by a reduction in cardiac output and was accompanied by an increase in systemic and pulmonary vascular resistance. The plasma concentrations of 6-ketoprostaglandin F1 alpha and thromboxane B2 also increased. These changes were markedly attenuated by aspirin. Both atrial pressures decreased during the second phase, thereby indicating that the PAF-induced reduction in cardiac output was related to a hindrance in venous return. The hematocrit increased, and aspirin did not affect this change. The extravasation of plasma probably plays a minor role, whereas venodilation would be the primary mechanism of the second-phase hypotension. S-1452, a prostaglandin H2/thromboxane A2 antagonist, abolished the PAF-induced pulmonary vasoconstriction but did not block the hypotensive action of PAF. OKY-046, a thromboxane A2 synthetase inhibitor, almost completely abolished the PAF-induced pulmonary vasoconstriction and the increase in plasma thromboxane B2 level, whereas it potentiated the hypotension and the increase in the plasma concentrations of prostaglandins; aspirin abolished this potentiation. These results suggest that PAF causes hypotension by two different mechanisms: 1) dilatation of resistance vessels independent of prostaglandins and 2) reduction of venous return due to venodilation, as mediated by prostaglandin(s).