The efficacy of intermittent pneumatic compression in the prevention of proximal deep venous thrombosis was determined in a consecutive series of patients who had primary or revision total hip replacement. Two hundred and eighty-nine patients (242 primary and eighty revision hip replacements) who were at least forty years old were managed intraoperatively and postoperatively with intermittent pneumatic compression with use of thigh-high sequential compression sleeves and thigh-high elastic compression stockings as the only form of prophylaxis. Venous ultrasonography of the ipsilateral lower extremity was performed, at an average of five days postoperatively, to determine the presence of a thrombus in the femoral and popliteal veins.The prevalence of proximal deep venous thrombosis was 6 per cent (twenty of 322 procedures), and no patient had a clinically detected pulmonary embolism. The prevalence of a proximal thrombus was 4 per cent (ten of 233 procedures) when a regional anesthetic had been used and 11 per cent (ten of eighty-nine procedures) when a general anesthetic had been used; this difference was significant (p = 0.02). The prevalence of a proximal thrombus was 3 per cent (eight of 245 procedures) for patients who were less than seventy-five years old and 16 per cent (twelve of seventy-seven procedures) for patients who were seventy-five years old or more (p < 0.0001). No patient had a major bleeding complication.Twenty-nine other patients who had an additional risk factor for thrombosis, such as a history of deep venous thrombosis, were managed with intermittent pneumatic compression and low-dose warfarin. The prevalence of proximal deep venous thrombosis in this group of patients was 19 per cent (six of thirty-one procedures).Intraoperative and postoperative intermittent pneumatic compression, combined with the use of a regional anesthetic, was found to be highly effective in the prevention of proximal deep venous thrombosis after total hip replacement. Despite the limitations inherent in comparisons among series, it is noteworthy that the prevalence of proximal thrombosis in our series was similar to the reported prevalences in several large series of patients who had a total hip replacement and were managed with either low-dose warfarin or low-molecular-weight heparin. However, there was no risk of major postoperative bleeding in the current study.