High mixed venous oxygen saturation (SVO2) values due to limited oxygen extraction capacities are reported in some patients with severe liver disease. The aim of this study was to evaluate the usefulness of SVO2 monitoring to assess adequate oxygen supply in such patients scheduled for liver transplantation. Nineteen patients with oxygen extraction ratio below 12% were analysed and compared to 20 patients with a pre-operative ratio over 17%. The two groups were comparable with regard to initial pathology, preload and haemoglobin levels. SVO2 values measured discontinuously by co-oximetry were unaffected by the first part of surgery and the clamping period in patients with low oxygen extraction ratio. In these patients, SVO2 was never correlated to oxygen supply during the whole procedure whereas good correlation was noted before and after unclamping in the other group. Tissue hypoxia detected by a dependent oxygen consumption-oxygen supply relationship occurred at clamping and unclamping in patients with initial low oxygen extraction capacities. It is concluded that expensive SVO2 continuous monitoring may not be effective in reflecting changes in oxygen supply in anaesthetized patients with initial severely impaired oxygen extraction capacity.