The choice of noninvasive tests used in primary prevention of cardiovascular diseases must be based on medical evidence. The aim of this study was to assess the additional prognostic value, over conventional risk factors, of physical examination, exercise testing, and arterial ultrasonography, in predicting a first coronary event. A prospective cohort study was conducted between 1996 and 2004 (n = 2,709), with follow-up in 2006 (response rate 96.6%). Participants had no history or symptoms of cardiovascular disease and had a standardized physical examination, a cardiac exercise testing, and carotid and femoral ultrasonography at baseline. Incident cases of definite coronary events were recorded during follow-up. Over the Framingham risk score, femoral bruit, positive exercise test, intima-media thickness >0.63 mm, and a femoral plaque provided significant additional information to the prediction model. The addition of the exercise test to the traditional risk factors, then the intima-media thickness and lastly the presence of femoral plaques, produces incremental increases in the area under the receiver operating characteristic curve (0.73-0.78, P = .02) and about a 50% increase in the positive predictive value (15.8%-31.4%), with no effect on the negative predictive value (96.4%-96.9%). Physical examination, exercise testing, and arterial ultrasonography provide incremental information on the risk of coronary event in asymptomatic adults. Exercise testing and femoral ultrasonography also improve the accuracy of the risk stratification.