We studied the diagnostic value of exercise electrocardiographic (ECG) testing in 191 patients who were completely asymptomatic 6 months after a successful percutaneous transluminal coronary angioplasty procedure. With > 70%- and > 50%-diameter stenosis at follow-up as restenosis criteria, the sensitivities of exercise ECG testing were 29% and 21%; the specificities 89% and 91%; the positive predictive values 20% and 52%; the negative predictive values 93% and 70%; the accuracies 83% and 68%; and the risk ratios 2.8 and 1.7, for prevalences of 9% and 33%, respectively. There were no significant differences in the diagnostic value of exercise ECG testing between men and women, patients receiving or not receiving beta-blocking agents, and the presence or absence of pathologic Q waves. Significant differences in systolic blood pressure and the rate-pressure product at peak exercise were found between patients with and without restenosis. For individual patients, however, no practical conclusions can be drawn from these values. In conclusion, the diagnostic value of exercise ECG testing for silent restenosis is low, and supplementation with other techniques seems to be warranted.