We sought in this study to elucidate whether carotid artery disease detected by ultrasonography can be a clinically useful marker for the presence of severe coronary artery disease (CAD) in patients evaluated for chest pain. Duplex ultrasonography and quantitative coronary angiography were used to assess carotid and coronary artery atherosclerosis in 225 consecutive patients (mean age, 58+/-9 years) with chest pain referred for cardiac catheterization. CAD was present in 197 patients (88%). Fifty-seven patients (25%) had 1-vessel disease, 52 (23%) had 2-vessel disease, 53 (24%) had 3-vessel disease, and 35 (16%) had left main stem CAD (LMS-CAD). The incidence of severe CAD (3-vessel disease or LMS-CAD) was 24% and 63% in the normal and impaired ejection fraction (EF) subgroups, respectively (P<0.005). Carotid disease (lumen diameter stenosis of >/=50%) was present in 5.3%, 13.5%, 24.5%, and 40% of patients with 1-, 2-, and 3-vessel disease and LMS-CAD, respectively. Moreover, the incidence of carotid disease in patients with severe CAD was 31% in the entire study population and 46% and 5% in the subgroups with impaired and normal EF, respectively (P<0.005). In the entire study population, the presence of severe CAD was determined by age, male sex, and carotid disease; in the impaired EF group by age and carotid disease; and in the normal EF group only by age. Carotid disease has a high negative (92%) and a high positive (91%) predictive value for the presence of severe CAD in the subgroup with normal and impaired EF, respectively. In patients evaluated for chest pain, carotid disease is significantly correlated with severe CAD. Furthermore, in patients with impaired left ventricular systolic performance the presence of carotid disease reflects the presence of severe CAD, while in patients with normal EF the absence of carotid disease reflects the absence of severe CAD.