Abstract The role of arachidonic acid metabolites in oleic acid-induced lung injury in anesthetized dogs was investigated. Oleic acid was administered as a bolus injection into the pulmonary artery in the following dose sequence: 0.05, 0.10, and 0.20 ml/kg at 30, 60, and 120 min, respectively, after either indomethacin (10 mg/kg iv) or vehicle. A control group ( n = 3) received normal saline instead of oleic acid. Measurements of hemodynamic parameters, mean systemic (MAP), pulmonary capillary wedge, and pulmonary artery pressures (PAP), cardiac output, arterial blood gases, extravascular lung waters (EVLW) by thermaldye double indicator dilution techniques and plasma immunoreactive thromboxane B 2 (iT×B 2), by radioimmunoassay were obtained at zero time (baseline) and 20 min following each oleic acid injection. A new noninvasive technique was employed to measure pulmonary capillary protein leak by the scintigraphic analysis of intravenously administered technetium-99m radiolabeled human serum albumin ( 99mTc-HSA) in the cardiac and lung regions. Oleic acid injection caused a significant dose related fall in MAP ( P < 0.0002), arterial pO 2 ( P < 0.0001), and cardiac output ( P < 0.001), and increases in EVLW ( P < 0.003) and plasma iT×B 2 ( P < 0.02) in the vehicle pretreated animals, while mean PAP remained unchanged. In contrast, in the indomethacin pretreated dogs, MAP, EVLW, cardiac output, and plasma iT×B 2 levels did not change from baseline values and there was an increase in mean PAP. Pulmonary vascular resistance was significantly elevated ( P < 0.05) in both groups. Compared to the control group, significant pulmonary capillary protein leaks occurred at the 0.10 ( P < 0.05) and 0.20 ( P < 0.01) ml/kg oleic acid doses in the vehicle pretreated dogs. Indomethacin failed to prevent the pulmonary capillary protein leak. There was a significant ( P < 0.02) correlation ( r = 0.72) observed between plasma iT×B 2 levels and maximal percentage protein leaks for the combined indomethacin and vehicle pretreated dogs. It was concluded that thromboxanes may serve as markers of oleic acid-induced lung injury but do not appear to mediate the resultant changes in pulmonary microvascular permeability.