One hundred and ten female patients, over the age of 60, with intertrochanteric or transcervical fractures were included in a controlled, randomized, clinical trial investigating the value of low dosage heparin in the prophylaxis of deep vein thrombosis. There were 50 completed pairs. Eight (16 per cent) deep vein thromboses occurred in the heparinized group compared with 23 (46 per cent) deep vein thromboses in the control group. The incidence of pulmonary embolism was also reduced. The diagnosis of deep vein thrombosis was made on clinical grounds, supplemented by phlebography and autopsy. There was no difference in the wound hematoma or infection rate. The heparin was commenced on admission to hospital and it is suggested that in this group of patients low dosage heparin prophylaxis should start on admission and not wait until surgery.