The purpose of this study was to determine if there are any morphologic characteristics of lesions that renarrow (that is, restenotic lesions) following successful coronary balloon angioplasty that are different from their appearance pretreatment or from the appearance of nonrestenotic lesions that might provide some new insight into the restenosis phenomenon. The study population consisted of 653 patients (778 lesions) with 6 months of angiographic follow-up (94% angiographic follow-up rate) who were participating in the Multicenter European Research trial with Cilazapril after Angioplasty to prevent Transluminal coronary Obstruction and Restenosis (MERCATOR) study. Detailed quantitative angiographic measurements, including the mean diameter of the vessel segment (in millimeters) that was subjected to balloon dilation, were performed preangioplasty, postangioplasty, and at follow-up using the cardiovascular angiographic analysis system to provide some objective measurement of the actual extent of luminal changes in the months following coronary balloon angioplasty. Two different approaches for restenosis were used: (1) static criterion of > 50% diameter stenosis at follow-up and (2) dynamic criteria of > or = 0.40 or > or = 0.72 mm change in minimal lumen diameter between postangioplasty and follow-up. Both approaches identified more severe stenosis to be a typical feature for restenotic lesions before angioplasty compared with nonrestenotic lesions. No differences were observed in lesion length, balloon-inflated vessel segment, or roughness index before angioplasty between the groups. Conflicting data were found for the amount of atherosclerotic plaque, symmetry index, and curvature index. The restenotic lesion at follow-up compared with its initial appearance gave conflicting results depending on which approach was used. The dynamic criteria illustrate that the reference diameter and the mean diameter of the entire segment dilated are reduced during follow-up. Two messages emerge from the study: (1) the restenosis process clearly involves the apparent normal vessel wall adjacent to the actual lesion, probably in response to the unavoidable injury caused by balloon dilatation and (2) the use of percentage diameter stenosis measurements depending on the assumptions of normality for a reference segment will therefore underestimate the true extent of the restenosis process and should be replaced in clinical angiographic studies by absolute luminal measurements.