Many angiographic definitions have been proposed to define restenosis after coronary angioplasty. The utility of each remains poorly defined. The aims of this study were: a) to analyse groups of patients defined by each of three criteria: > 50% stenosis (definition 1), loss > or = 50% of initial gain in diameter (definition 2), loss > or = 0.52 mm of minimal luminal diameter based on the variability of the angiographic measurement (definition 3) and, b) to compare the immediate attitude of the interventional cardiologist with the deferred quantitative angiographic analysis. The angiographic follow-up included 89 patients. The angiographic restenosis rate was 37% (definition 1), 48% (definition 2) and 43% (definition 3). Restenosis as defined by criterion 1 was associated with the greatest degree of postangioplasty residual stenosis (p = 0.02) whereas, with criteria 2 and 3, it was associated with less severe residual stenosis (p = 0.03 and p = 0.007). Definition 2 and 3 are the most similar and definitions 1 and 3 the most complementary. The sensitivity, specificity positive and negative predictive values for recurrence of angina with respect to angiographic restenosis (definition 1) were respectively 63.6%, 77.8%, 63.6%, and 77.8% and are not significantly improved by associated analysis of exercise testing. Discordances between the decision of the interventional cardiologist and the results of quantitative angiography (definition 1) were noted in 12.4% of the stenosis studied, there measuring 44 to 64%. The judgement of the cathetiser of these intermediary stenoses was essentially influenced by the recurrence of angina during follow-up.(ABSTRACT TRUNCATED AT 250 WORDS)