Carotid ultrasound is a useful tool to detect plaque and measure carotid intimal-medial thickness (CIMT), which help identify asymptomatic patients at risk of cardiovascular events. However, the role of carotid ultrasound in stratifying risk in symptomatic patients has not been studied. We prospectively examined the hypothesis that in symptomatic patients undergoing exercise (or pharmacologic) stress testing (EST), CIMT and carotid plaque will improve the ability to identify patients with > or =50% coronary diameter stenosis. Mean and maximal CIMTs were measured in 50 symptomatic patients without known coronary artery disease (CAD) who were scheduled to undergo diagnostic angiography after EST with imaging. Pretest probability of CAD was calculated for each patient based on age, gender, and nature of symptoms. Fifty percent of patients were men, 32% were diabetic, 70% had hypertension, and 58% had dyslipidemia. CAD > or =50% was associated with CIMT in the >75th percentile (74% vs 44%, p = 0.047) and with presence of carotid plaque (96% vs 59%, p = 0.003). Of patients with an equivocal/negative EST result (n = 38), those with carotid plaque had a 46% (13 of 28) prevalence of significant CAD; in the smaller subset of patients without carotid plaque (n = 10), none had luminal stenosis > or =50% (p = 0.007). Pretest probability of CAD, presence of carotid plaque, moderately/severely abnormal EST result, and hypertension were independent predictors of significant CAD. In conclusion, carotid ultrasound appears to offer incremental value to EST in predicting presence or absence of > or =50% coronary diameter stenosis in symptomatic patients with equivocal/negative EST result.