In order to confirm a complete ischemia model, 1-hour warm hepatic ischemia by hepatic vascular exclusion (HVE) was studied in dogs, in comparison with that by inflow occlusion (IOC) only. The splanchnic venous bed and/or infrahepatic inferior vena cava were decompressed by a centripetal pump-driven venovenous bypass. Indocyanine green retention test revealed no hepatic blood flow in the HVE model during ischemia, while hepatic blood perfusion was still present in the IOC model. All 5 of the IOC dogs survived more than 7 days after revascularization, while 4 of the 5 HVE dogs died within 9 h. After the induction of hepatic ischemia, lactate increased in both HVE and IOC dogs. After revascularization, transaminases and guanase were elevated, the arterial ketone body ratio (acetoacetate/3-hydroxybutyrate) decreased and the serum lactate accumulated more in HVE dogs than in IOC dogs. The hepatic redox state of IOC dogs was significantly decreased by additional clamping of the inferior vena cava. It is concluded that the HVE model with a pump-driven active bypass provides complete and stable hepatic ischemia, resulting in greater deterioration of hepatic cellular functions; hence it is more suitable as a model of complete hepatic ischemia than the IOC one.