on behalf of the CARPORT Study Group Because many ongoing clinical restenosis prevention trials are using quantitative angiography to assess whether a drug is capable of reducing the amount of intimal hyperplasia, quantitative angiographic risk factors for angiographic luminal narrowing after balloon angioplasty were determined, including stretch and elastic recoil at the dilatation site. Quantitative analysis was performed on 666 lesions in 575 patients during angioplasty and at 6-month follow-up. Stretch was defined as balloon diameter minus minimal luminal diameter (MLD) before angioplasty/reference diameter, and recoil as balloon diameter minus MLD after angioplasty/reference diameter. Multivariate analysis was used to yield independent risk factors for luminal narrowing at follow-up. Predictors of absolute change in MLD were (1) relative gain at angioplasty (gain in millimeters normalized for reference diameter) and (2) lesion length. To allow risk stratification, logistic regression analysis was applied using the decrease in MLD as a binary outcome variable. A decrease in MLD at follow-up of 20.72 mm was considered significant. Variables retained in the model were: relative gain >0.3 mm (rate ratio 2.9), relative gain 0.2 to 0.3 (rate ratio 2.1), stenosis length 16.8 (rate ratio 1.7), and thrombus after angioplasty (rate ratio 2.6). Although stretch was significantly related to luminal narrowing at univariate analysis, it was not retained in the multivariate models.A large gain in lumen diameter at angioplasty, dilation of long lesions, and angiographically determined thrombus after angioplasty were found to be accompanied by more severe luminal narrowing at follow-up. (Am J Cardiol 1992;69:564-591)