As O2 delivery falls, tissues must extract increasing amounts of O2 from blood to maintain a normal O2 consumption. Below a critical delivery threshold, increases in O2 extraction cannot compensate for the falling delivery, and O2 uptake falls in a supply-dependent fashion. Numerous studies have identified a critical delivery in whole animals, but the regional contributions to the critical O2 delivery are less fully understood. In the present study, we explored the limits of O2 extraction in the isolated liver, seeking to determine 1) the normal relationship between O2 consumption and delivery in the liver and 2) the relationship of hepatic lactate extraction to the drop in hepatic O2 consumption at low O2 deliveries. To answer these questions, using support dogs as a source for oxygenated metabolically stable blood, we studied eight pump-perfused canine livers. By lowering the blood flow in a model of stagnant hypoxia, we explored the relationship between O2 consumption and delivery over the entire physiological range of O2 delivery. The critical O2 delivery was 28 +/- 5 (SD) ml.kg-1.min-1; the livers extracted 68 +/- 9% of the delivered O2 before reaching supply dependence. This suggests that the liver has an O2 extraction capacity quite similar to the body as a whole and not different from other tissues that have been isolated. At high blood flows, the livers extracted approximately 10% of the lactate delivered by the blood, but the arteriovenous lactate differences were small. At low blood flows, however, the livers changed from lactate consumption to production. The O2 delivery coinciding with the dropoff in lactate extraction did not differ significantly from the critical O2 delivery. We conclude that reductions in lactate uptake by the liver do not precede the transition to O2 supply dependence.