Levels of plasma crosslinked fibrin derivatives, a sensitive and direct marker of the lysis of intravascular crosslinked fibrin, were measured serially in 135 patients undergoing major abdominal surgery to determine their behavior and their use as a screening test for postoperative venous thrombosis. Preoperative levels and levels on the first postoperative day were significantly higher by both enzyme immunoassay and latex assay in 31 patients who developed venous thrombosis (positive venography) than in 104 patients who did not (negative125I fibrinogen leg scan). Preoperative XLFDP levels 400 ng/ml (enzyme immunoassay) had a sensitivity to the diagnosis of postoperative venous thrombosis of 58%, specificity 74%, positive predictive value 41% and negative predictive value of 85%. The sensitivity of XLFDP levels over 1200 ng/ml on the first postoperative day was 65%, specificity 73%, positive predictive value 38% and negative predictive value 89%. These cutoff values were chosen (high negative predictive value) to allow identification of patients who were unlikely to have venous thrombosis. Measurement of plasma XLFDP, a simple inexpensive test, could be used as a screen to select patients for surveillance procedures (IPG or duplex ultrasonography). A substantial increase in XLFDP levels (> 500 ng/ml) occurred in virtually all patients, suggesting that fibrinolysis is not ‘shutdown’ postoperatively and that these assays reflect lytic activity at the fibrin surface more accurately than do measurements of plasminogen activators and their inhibitors.