Abstract. Generation of reactive oxygen metabolites, thromboxane increases, and vasoconstriction have been implicated in the pathogenesis of acute edematous lung injury, such as that seen in patients with the Adult Respiratory Distress Syndrome (ARDS), but their interactions are unknown. We hypothesized that reactive 02 products would stimulate arachidonic acid metabolism in lungs and that vasoactive products of arachidonate, such as the potent vasoconstrictor thromboxane A2, might then mediate 02-metabolite-induced pulmonary vasoconstriction. We found that 02 metabolites generated by injection of purine plus xanthine oxidase caused increases in mean pulmonary artery perfusion pressures (27±4 mmHg) in isolated perfused lungs. In addition, purine plus xanthine oxidase also caused 30-fold increases in perfusate levels ofthromboxane B2 (the stable metabolite ofthromboxane A2) compared with only twofold increases in 6-keto-PGFia (the stable metabolite of prostacyclin). Moreover, prior addition of catalase inhibited both vasoconstriction and the thromboxane B2 production seen in isolated lungs following injection of purine plus xanthine oxidase. Similarly, pretreatment with cyclooxygenase inhibitors, either aspirin or indomethacin, also completely blocked thromboxane generation and markedly attenuated pressor responses usually seen after purine plus xanthine oxidase (increase in mean pulmonary artery perfusion pressures, 4.4±1.5 mmHg). Furthermore, imDr. Tate is a recipient of a Clinician-Scientist Award from the American Heart Association. Address correspondence and reprint requests to Dr.Tate, Pulmonary Division, Department of Medicine, Box 4000, Denver General Hospital, Denver, CO 80204.Received for publication 6 July 1983 and in revised form 26 April 1984. idazole, a thromboxane synthetase inhibitor, also decreased 02-metabolite-induced thromboxane generation and vasoconstriction. These results suggested that thromboxane generation might participate in 02-metaboliteinduced vasoconstriction. However, since a significant correlation between thromboxane levels and the degree ofvasoconstriction could not be demonstrated, and since addition of superoxide dismutase reduced thromboxane generation but did not affect the intensity of vasoconstriction, it is possible that thromboxane is not the only vasoactive mediator in this model. We conclude that exposing lungs to 02 metabolites results in thromboxane generation and that thromboxane is a major mediator of oxidant-induced vasoconstriction.