The diagnosis of deep venous thrombosis in the lower extremity has usually been thought to mandate the initiation of anticoagulation therapy. However, there has been evidence to show that this approach may not be necessary for deep venous thrombosis at or distal to the knee, and that the risks of treatment may be greater than those of the thrombosis itself. One hundred and seventy-five patients who underwent total knee replacement were examined postoperatively by venography, plethysmography, fibrinogen scans, and ventilation-perfusion lung scans. We found that 126 (72 per cent) of them had small or large clots in the calf, and that only forty-nine (28 per cent) had no thrombi. Seventy-one (41 per cent) of the patients had small thrombi in the calf and fifty-five (31 per cent) had large thrombi in the calf. Six patients had thrombi in the thigh, all of which were associated with large thrombi in the calf. In only two patients, however, did clinically recognized pulmonary emboli develop, one in the group of patients without known thrombi and the other in a patient with a large iliofemoral thrombus. The ventilation-perfusion scans showed six asymptomatic pulmonary emboli that were not associated with the presence of either large or small thrombi in the calf. The fibrinogen scans that were done in the postoperative period were capable of revealing large but not small thrombi in the calf. The preoperative plethysmography did not aid in determining in which patients a large thrombus of the calf or thigh was likely to develop.(ABSTRACT TRUNCATED AT 250 WORDS)