We performed a double-blind randomized trial comparing high doses of subcutaneous heparin (12,500 units every 12 hours) with low doses (5000 units every 12 hours) for 10 days in the prevention of left ventricular mural thrombosis in 221 patients with acute anterior myocardial infarction. Left ventricular mural thrombosis was observed by two-dimensional echocardiography on the 10th day after infarction in 10 of 95 patients (11 percent) in the high-dose group and in 28 of 88 patients (32 percent) in the low-dose group (P = 0.0004). One patient in the high-dose group and four in the low-dose group had nonhemorrhagic strokes (P = 0.17). One patient in the low-dose group had a fatal pulmonary embolism. There was no difference in the frequency of hemorrhagic complications, which occurred in six patients in the high-dose group and four in the low-dose group. The mean (+/- SEM) plasma heparin concentration was 0.18 +/- 0.017 U per milliliter in the high-dose group and 0.01 +/- 0.005 U per milliliter in the low-dose group (P less than 0.0001). In the high-dose group, the mean plasma heparin concentration was 0.10 +/- 0.029 U per milliliter among patients with abnormal two-dimensional echocardiograms, as compared with 0.19 +/- 0.019 U per milliliter among patients with normal echocardiograms (P = 0.01). We conclude that heparin administered subcutaneously in a dosage of 12,500 units every 12 hours to patients with acute anterior transmural myocardial infarction is more effective than a lower dosage (5000 units every 12 hours) in preventing left ventricular mural thrombosis.