The diagnosis of pulmonary thromboembolism is frequently based on ventilation-perfusion scintigraphy and ascending lower limb venography when pulmonary angiography is not available. The aim of this study is to compare color Doppler ultrasound against ascending venography in the evaluation of the lower limb deep vein system in patients with clinical suspicion of pulmonary embolism, with special attention to calf veins. We prospectively studied 30 patients with clinical suspicion of pulmonary embolism in whom a color Doppler ultrasound and venogram were performed with no more than a 3-h interval between both procedures. The diagnostic criteria was that of loss of venous compressibility. The 'color' ability was used to identify artery from vein. Out of 15 patients in whom a venogram proved positive (50%), 9 had isolated calf vein thrombosis (60%). In 5 patients, the color Doppler ultrasound of the calf was considered inconclusive. Overall sensitivity was 53%, specificity 100%, positive predictive value 100%, and negative predictive value 68%. In the femoropopliteal system, sensitivity was 83% and specificity 100%. Considering all patients, sensitivity in the calf system was 40%. Excluding the 5 patients who were difficult to assess, sensitivity increased to 60%. In conclusion, color Doppler ultrasound is not as sensitive as venography in dealing with patients with clinical suspicion of pulmonary embolism, due to its low sensitivity in the calf system when distal thrombi need to be excluded. However, a reasonable alternative is to begin by performing a compression ultrasonography of the femoropopliteal system. Color Doppler ultrasonography of the calf system represents a rarely sensitive and arduous task and does not seem justifiable in this type of patient.(ABSTRACT TRUNCATED AT 250 WORDS)