Between November 1991 and June 1992, 50 patients suspected of pulmonary embolism (SP > 20%) underwent Doppler echocardiography, venous Doppler ultrasonography and pulmonary angiography. Pulmonary embolism was confirmed by pulmonary angiography in all patients but 3 (2 pts: mean pulmonary pressure > 50 mmHg and 1 pt: mobile thrombus between the infundibulum and the main pulmonary artery). Two groups were identified on the basis of Miller's index: Group 1: "non-massive" pulmonary embolism, Miller < 60% (n = 18); Group 2: "massive" pulmonary embolism, Miller > or = 60% (n = 29). The patient with thrombus in the main pulmonary artery and the two with high pulmonary pressures were included in Group 2. Venous Doppler ultrasonography was performed in 96% (n = 48) of patients, including 90% within the first 24 hours. No distinction could be drawn between the two groups on the basis of venous Doppler ultrasonography findings. A majority of patients had thrombosis of main collecting vessels (Group 1 = 75%, Group 2 = 78%) and 10% of patients had no venous thrombosis of the lower limbs. Doppler echocardiography was performed in all patients, including 94% (n = 47) within the first 24 hours. Dilatation of the left ventricle as well as analysis of septal contraction was evaluable in all patients except one of Group 2, because of poor technical quality and of artificial pacing. A RV/LV ratio > 0.60 was found in 97% (30/31) of patients of Group 2 as compared with 39% (7/18) in Group 1.(ABSTRACT TRUNCATED AT 250 WORDS)