Seventy-two patients undergoing elective total hip replacement were studied with bilateral venography, pre-operative and post-operative lung scans. Twelve belonged to a control group that received placebo injections and 60 patients to treatment groups that received low molecular weight heparin. The incidence of deep venous thrombosis was 11 (92%) of 12 patients in the control group and 18 (30%) of 60 patients in the treatment group (X2: P < 0.001). The incidence of pulmonary embolism (new unmatched perfusion defects) was five (42%) of the 12 patients in the control group and five (8.3%) of the 60 patients in the treatment groups (X2: P < 0.002). The incidence of pulmonary embolism (new unmatched perfusion defects) was eight (27.5%) of 29 patients with deep venous thrombosis and two (4.6%) of 43 without deep venous thrombosis (X2: P < 0.02). Of the ten patients who had pulmonary embolism according to this study's criteria (one or more new defect on perfusion, unmatched on the ventilation scan), eight would have been classified as high probability by the Biello criteria, and only five by the PIOPED criteria, if the pre-operative scans were not available. We conclude that having a pre-operative lung scan improves ability to interpret the postoperative lung scans in high risk patients.