Twenty-eight consecutive patients with acute spinal cord injury were randomised to one of two regimens for the prevention of deep vein thrombosis (DVT): external pneumatic calf compression (EPCC, 15), or EPCC combined with aspirin, 300 mg bid, and dipyridamole, 75 mg tid (ASA/dip, 13). DVT, detected by the 125I-fibrinogen test and impedence plethysmography, was confirmed by contrast venography. The incidence of DVT in the total group was 33 per cent, significantly less than the 78 per cent observed in 37 untreated patients studied previously (p less than 0.001). DVT developed in six of 15 patients receiving only EPCC, and three of 12 on ASA/dip as well as EPCC (p less than 0.1). The untreated patients studied earlier had significantly increased circulating platelet aggregates, increased platelet affinity for collagen, and elevated factor VIII activities, which generally coincided with the development of DVT (usually 7-9 days after injury). Prophylaxis partially prevented these coagulation abnormalities and delayed the onset of thrombosis. While the differences in the haemostatic parameters between those not treated with ASA/dip and those receiving these agents were not statistically significant (except for the platelet-collagen affinity), there was a trend toward less elevated values in the drug-treated patients. We conclude that spinal cord injury patients are at extreme risk for DVT, and have abnormal platelet and factor VIII activities. Prophylaxis with EPCC significantly and safely reduces the risk of DVT in these patients.