A consecutive series of 192 patients (121 men and 71 women, mean age 59 years, range 28 to 82) with isolated, severe valvular aortic stenosis was with isolated, severe valvular aortic stenosis was analyzed retrospectively to determine the relation of angina pectoris and coronary risk factors to angiographically significant coronary artery disease (CAD). Significant CAD (diameter reduction greater than or equal to 50%) was found in 47 patients (24%). Angina was present in 83% of them, but it was also found in 61% of the non-CAD patients. This symptom had as a result a low positive predictive value (31%). Of the patients without angina (n = 65) 12% had significant CAD. The negative predictive value of angina alone was thus 88%. By using multivariate logistic regression, a risk score could be calculated based on angina, age and sex, which increased the negative predictive value to 95%. It was concluded that coronary arteriography can only be omitted in severe aortic valvular stenosis, when patients have no angina and when they are less than 40 years of age for men and less than 50 years for women. For all other cases, coronary arteriography should be recommended.