Intra-arterial injections of platelet-activating factor (Paf-acether, 10-300 ng) to the perfused guinea-pig lung induced a dose-related bronchoconstriction, followed by contraction of the rat aorta superfused with the lung effluent, indicating the release of thromboxane A2 (TXA2) activity. These effects were matched with injections of bradykinin (Bk) at 100-1000 ng, leukotriene C4(LTC4) at 10-300 ng or arachidonic acid (AA) at 30-300 micrograms. Repeated doses of Paf-acether led to a specific desensitization of the release of TXA2, under conditions where Bk, LTC4 and arachidonic acid retained their ability to release TXA2. Bronchoconstriction and the release of TXA2 induced by Paf-acether were suppressed when the lungs were perfused with acetylsalicylic acid, but not with salicylic acid. The phospholipase A2 inhibitor, p-bromophenacyl bromide suppressed the release of TXA2 by Bk, but did not interfere with its formation from AA, nor with its release with Paf-acether and LTC4. The lipoxygenase inhibitor, nordihydroguaiaretic acid, inhibited to a similar extent the release of TXA2 by Bk, LTC4 and Paf-acether but also reduced directly the formation of TXA2 from arachidonic acid, invalidating its use as a specific antilipoxygenase agent. The leukotriene C4/D4 antagonist, FPL 55712, suppressed the TXA2 releasing effects of LTC4, and was completely inactive against Paf-acether, Bk or arachidonic acid. The aerosol of Paf-acether was tested in the anaesthetized guinea-pig and resulted in bronchoconstriction, unaccompanied by thrombocytopenia. Unlike bronchoconstriction induced by intravenous Paf-acether, which is refractory to cyclo-oxygenase inhibitors, the effects of the aerosol were suppressed by aspirin. Platelet depletion, which blocks the intravenous effects of Paf-acether, failed to interfere with those of the aerosol. Paf-acether induced a marked contraction of the superfused guinea-pig isolated parenchyma lung strip, which was followed by total and irreversible desensitization to itself. The contractile effect was not inhibited by aspirin or indomethacin, atropine, mepyramine, methysergide, phenoxybenzamine or propranolol, indicating that cyclo-oxygenase products, cholinergic stimuli, histamine, 5-hydroxytryptamine and catecholamine mechanisms are not involved. Our results indicate that Paf-acether interacts with pulmonary sites distinct from those for Bk, LTC4 or AA, since no cross-desensitization between Paf-acether and the other agonists was noted, p-bromophenacyl bromide inhibited Bk only and FPL 55712 inhibited only LTC4. The phospholipase A2 involved with the release of the arachidonate needed for the formation of TXA2 by Paf-acether or LTC4-stimulated lungs may differ from the enzyme accounting for its formation by Bk. The cellular sites with which Paf-acether interacts may also be distinct and less readily accessible to p-bromophenacyl bromide.