We hypothesized that the extent of acute lung injury (ALI) caused by lipopolysaccharide (LPS) is modified with its initial passage through the liver. We tested this hypothesis by administering LPS, 5 mg/kg, or saline to 120 male Wistar rats via the portal vein (PV) or the inferior vena cava (IVC) over 1 h. Four experimental groups of rats were administered saline into the PV, saline into the IVC, LPS into the PV (LPS-PV group), and LPS into the IVC (LPS-IVC group), respectively. At 15 and 30 min after onset of 51Chromium-LPS infusion, the gamma counts in the liver were higher in the LPS-PV group than that in the LPS-IVC group. The ratio of 125Iodine-albumin counts in lung tissue to that in plasma per unit of weight (as an assessment of pulmonary microvascular permeability) at 240 min after onset of LPS stimulation, the accumulation of polymorphonuclear cell (assessed by myeloperoxidase activity) and the concentration of tumor necrosis factor alpha in the lung at 60 and 240 min after onset of LPS infusion, were higher in the LPS-IVC group than in the LPS-PV group. Significant differences in several factors indicative of inflammation and in the extent of LPS-induced ALI were observed after the onset of LPS infusion, depending on whether it was delivered via the PV or the IVC. These observations suggest that the entrapping of LPS during its initial passage through the hepatic circulation may attenuate LPS-induced ALI within 4 h of initiation of LPS stimulation.