To evaluate the clinical significance of observed R wave amplitude changes in exercise-induced supraventricular extrasystoles in comparison to the preceding sinus beat, 94 patients catheterized for possible coronary artery disease (CAD) were studied. Significant CAD was documented in 63 patients--34 with myocardial infarction (group A1) and 29 without (group A2)--whereas 31 patients had normal coronary arteries or coronary lesions less than 30% (group B). All patients underwent treadmill stress testing using the Bruce protocol within a month after cardiac catheterization. R wave amplitude increased or remain unchanged in extrasystole (R(x-s) > or = 0) in patients with CAD, while it decreased (negative R(x-s) in patients without significant CAD (P < .0001). In patients with CAD R(x-s) values were positively related to the number of obstructed coronary arteries (P < .01), while no significant difference was found between groups A1 and A2. The correlations of R wave amplitude changes in extrasystoles were significant with coronary obstruction score values (r = .82 and .85 in groups A1 and A2, respectively) and with left ventricular ejection fraction values (r = -.88, -.86 and -.90 in groups A1, A2, and B, respectively). R(x-s) > or = 0 value had a sensitivity of 79% and a specificity of 90% for CAD detection, while sensitivity was higher (89%) and specificity was lower (57%) for the prediction of left ventricular dysfunction. It is concluded that R(x-s) > or = 0 value is indicative of CAD, multivessel disease, and poor left ventricular performance, while its negative value is combined with minimal or no CAD and normal ejection fraction values.