Background The procedural result is a major determinant of the incidence of 6-month target vessel revascularization (TVR) after successful coronary stenting. However, the prognostic implications of the different measures of the procedural result or procedural end points have not been directly compared. In this study, we sought to assess and compare the impact of achieving different procedural end points on the long-term (2-year) incidence of TVR. Methods and Results We studied 234 patients in whom 1 or 2 stents were successfully deployed and ultrasound imaging performed after angiographic optimization. End points included a visually estimated angiographic residual stenosis <10% and ultrasound stent-to-mean reference lumen area ≥80%. After 2 years, TVR was required in 48 (20.5%) patients. Qualitative predictors of TVR were vein graft lesions, 3-vessel disease, and baseline TIMI flow grade <3. Quantitatively, reference diameter by quantitative coronary angiography (QCA), final minimum lumen diameter (MLD) by QCA, and in-stent minimum lumen area (MLA) by ultrasound were predictive of TVR. Stent-to-reference ratios were not significantly predictive of TVR. By multivariable analysis, vein graft location and MLA by ultrasound were the only significant predictors of TVR (relative risk, 2.9 [1.5, 5.4] and 0.72 [0.6, 0.9], respectively). Receiver operator curves for MLD by QCA and MLA by ultrasound were similar in predicting TVR. Neither was significantly superior to reference vessel diameter. Conclusions Commonly used angiographic and ultrasound stent-to-reference ratios do not predict the incidence of TVR. Absolute measures of the lumen size (MLA by ultrasound and MLD by QCA) were the most important quantitative predictors of TVR within 2 years. This emphasizes the role of the vessel size as the limiting factor in determining the long-term outcome of coronary stenting. (Am Heart J 2001;141:823-31.)