The role of reactive oxygen species (ROS) in hepatic ischemia-reperfusion injury has been a controversial issue for more than 15 years. Evidence for xanthine oxidase, mitochondria, Kupffer cells, and/or neutrophils as source(s) of the postischemic oxidant stress has been presented in a number of different in vitro and in vivo models. The early focus was on intracellular sources of reactive oxygen formation; this was followed by emphasis on vascular oxidant stress. Recently, vascular reactive oxygen formation has been examined as a cause of intracellular oxidant stress in hepatocytes. The study by Kumamoto et al. 1 in the present issue of HEPATOLOGY emphasizes again the relevance of intracellular oxidant stress in the pathophysiology of reperfusion injury. Does this mean that we have gone full circle, that the early studies were correct all along?Substantial work with antioxidants (e.g., superoxide dismutase, catalase, deferoxamine, and vitamin E) initially made a strong case for the involvement of ROS in hepatic ischemiareperfusion injury (reviewed in Jaeschke 2 ). Studies with allopurinol pointed to xanthine oxidase as the major intracellular source of ROS formation in hepatocytes. 3 However, the xanthine oxidase hypothesis drew criticism for 3 reasons: First, the conversion of xanthine dehydrogenase to the oxidase occurred only after relatively prolonged ischemia. 4 Second, the extent of the postischemic intracellular oxidant stress was very limited at best. 5,6 Finally, superimposition of severe, chemically induced oxidant stress did not add to reperfusion injury. 6 These data threw into question not only the role of xanthine oxidase but also the pathophysiological relevance of intracellular oxidant stress during reperfusion. 2 At the same time, evidence for oxidant stress in the hepatic vasculature was emerging. 7,8 Kupffer cells were quickly identified as a major source of the early postischemic oxidant stress. Interventions that inactivated Kupffer cells attenuated the oxidant stress, 8 Kupffer cells from postischemic liver generated enhanced amounts of superoxide, 9 and ultrastructural evidence for activation of Kupffer cells during reperfusion was reported. 10 In addition, it was recognized that neutrophils accumulate in substantial numbers in the postischemic liver and contribute to the injury within 5 to 6 hours of reperfusion. 11 Showing the enhanced ROS formation of neutrophils isolated from the postischemic liver provided direct evidence for a neutrophil-induced oxidant stress. 12 The importance of this vascular source of ROS was shown by the functional inactivation of neutrophils by a Mac-1 (CD11b/CD18) antibody, which attenuated the oxidant stress and protected substantially against reperfusion injury. 13 At the same time, it was recognized that complement activation is a critical factor in activating Kupffer cells and neutrophils during reperfusion. 14 Despite a large number of follow-up studies supporting the concept of an inflammatory injury during reperfusion, individual reports, including th...