A range of clinical data was obtained from 124 patients about to undergo operation and several coagulation tests were performed. No patient received prophylaxis for deep vein thrombosis, and isotopic scanning after operation showed that 20 patients had developed thrombosis. a simiple prognostic index for predicting which patients would develop postoperative deep vein thrombosis was constructed using the clinical and coagulation data obtained before operation. The five variables with the best predictive power-euglobulin lysis time, age, presence of varicose veins, fibrin related antigen, and percentage overweight-produced an equation that identfied 95% of those who developed deep vein thrombosis and misallocated only 28% of those who did not develop thrombosis. In view of the complications that low-dose heparin and dextran can cause, giving prophylaxis to under a third of the patients who will not develop deep vein thrombosis is clearly better than giving it to all.