The scaffolding effect of stent implantation has the potential to alter vascular geometry and dimensions. The objective of this study was to determine the impact of intracoronary stent implantation on the reference vessel diameter and the consequences of this on the frequency of restenosis applying the binary definitions of restenosis based on percent diameter stenosis. Routine angiographic follow-up was performed in 79/80 consecutive patients who had a single elective Palmaz-Schatz stent implanted in denovo lesions in native coronary arteries 6.5 +/- 3.4 mo after the index procedure. Complete quantitative angiographic follow-up was available in 78 (98%). The mean reference vessel diameter was 2.9 +/- 0.6 mm preprocedure, increased to 3.1 +/- 0.5 mm immediately poststent implantation and was 2.6 +/- 0.6 mm at follow-up (F = 6.45, P = 0.0001, ANOVA for repeated measures). In view of the varying reference vessel diameter, percent diameter stenosis postangioplasty and at follow-up was determined by two methods: (1) automatically by the quantitative coronary angiographic analysis system and (2) by expressing the minimal luminal diameter postangioplasty and at follow-up as a function of the original preprocedural reference vessel diameter. The restenosis rate was significantly greater for all definitions of restenosis when the minimal luminal diameter was determined as a function of the original preprocedure reference vessel diameter (e.g., 34% vs. 18% for the > or = 50% criterion, P = 0.018). Stent implantation results in alterations in reference vessel diameter, which have important consequences for the frequency of restenosis presented as a binary variable based on percent diameter stenosis.