To determine the value of a 6-month exercise treadmill test for detecting restenosis after elective percutaneous transluminal coronary angioplasty (PTCA), 303 consecutive patients with successful PTCA and without a recent myocardial infarction were studied. Among the 228 patients without interval cardiac events, early repeat revascularization or contraindications to treadmill testing, 209 (92%) underwent follow-up angiography, and 200 also had a follow-up treadmill test and formed the study population. Restenosis (greater than or equal to 75% luminal diameter stenosis) occurred in 50 patients (25%). Five variables were individually associated with a higher risk of restenosis: recurrent angina (p = 0.0002), exercise-induced angina (p = 0.0001), a positive treadmill test (p = 0.008), more exercise ST deviation (p = 0.04) and a lower maximum exercise heart rate (p = 0.05). However, only exercise-induced angina (p = 0.002), recurrent angina (p = 0.01) and a positive treadmill test (p = 0.04) were independent predictors of restenosis. Using these 3 variables, patient subsets could be identified with restenosis rates ranging from 11 to 83%. The exercise treadmill test added independent information to symptom status about the risk of restenosis after elective PTCA. Nevertheless, 20% of patients with restenosis had neither recurrent angina nor exercise-induced ischemia at follow-up. For more accurate detection of restenosis, the exercise treadmill test must be supplemented by a more definitive test.