The optimal stimulus which produced the maximum increase in blood velocity in the femoral vein during compression of the lower limb with a sequential compression device (six chambers) was determined using Doppler ultrasound in 10 normal limbs. Pressures of 35, 30, and 20 mm Hg at the ankle, calf, and thigh, respectively, applied sequentially for 12 seconds produced a 240% increase in the peak blood velocity. Higher pressures did not increase velocity any further. A nonsequential device (one chamber) inflated at 35 mm Hg for 12 seconds produced only a180% increase in blood velocity. The efficacy of the sequential device to prevent deep venous thrombosis then was tested and compared with a single chamber device and small-dose subcutaneous heparin in a randomized, controlled clinical trial using the 125I-fibrinogen test. The results suggest that the sequential compression device is as effective as heparin during the period when it is used (the first 24 hours after operation) and more effective than a nonsequential device in preventing deep venous thrombosis proximal to the calf.