Tissue damage during surgery induces coagulation factors and activates platelets. Surgical pain may provoke release of catecholamines, leading to hypercoagulability. We have investigated the effect of surgical pain on blood coagulability and fibrinolysis in orthopaedic operations using tourniquets in 22 patients undergoing total knee replacement. Patients were allocated to one of two groups to receive extradural anaesthesia (EA; n = 11) or general anaesthesia (GA; n = 11). The EA group received lumbar extradural block with lidocaine. The GA group received only general anaesthesia, maintained with 1.5-2.5% sevoflurane and 66% nitrous oxide in oxygen. Using a thrombelastogram technique, blood coagulability and fibrinolysis were measured. Mean maximum amplitude (MA), which reflects coagulability, increased after tourniquet inflation (11%) in group GA whereas MA in group EA did not change. After tourniquet deflation, MA values in both GA and EA groups increased significantly (10% and 20%, respectively) (P < 0.05), and there was also a significant difference in MA between groups (P < 0.05). The fibrinolytic rate did not change in either group during tourniquet inflation, but increased significantly (160%) after tourniquet deflation. There was no significant difference in fibrinolytic rate between the groups. We conclude that the hypercoagulability seen in group GA could have been caused by surgical or tourniquet pain, or both, and that extradural anaesthesia is a useful technique to prevent hypercoagulability.