This study examined the optimal angiographic definition for long-term success after angioplasty and compared visual and quantitative angiographic measurements in assessing outcome. The National Heart, Lung, and Blood Institutes--Percutaneous Transluminal Coronary Angioplasty Registry prospectively followed 1,768 patients from 15 clinical centers. Symptom-free survival, defined as survival without angina, myocardial infarction, bypass surgery, or death, occurred in 59% of patients. In a subset of 393 patients, quantitative coronary angiography (QCA), done at a core angiographic laboratory, was compared with visual site readings. Although there was considerably more variability for visual readings, a site reading of a change in percent stenosis of >20% correlated highly with symptom-free survival (64.6% for patients who had all lesions successfully dilated, 48% for patients with partial success, and only 21% for patients without angiographic success; p < 0.001). Similar findings were seen for other angiographic definitions, but a change of > 20% was most discriminatory. In contrast, QCA readings had little or no predictive value. This study confirms that visual assessment of the immediate change in percent stenosis is predictive of a successful 1-year outcome. A change of greater than 20% is most discriminatory and should still be used to define angiographic success. QCA does not appear to be superior to visual assessment in predicting 1-year outcome.