A correlation of the angiographic evolution of coronary stenoses (stenosis diameter > or = 20%) with morphological stenosis parameters at baseline could help to identify the risk of progressive stenoses. Therefore, the data of the prospective INTACT study (International Nifedipine Trial on Antiatherosclerotic Therapy) were reviewed. In 348 patients with moderate coronary artery disease, standardized coronary angiograms were taken 3 years apart and were quantitatively analysed. Changes in the minimal diameter of the 1063 preexisting coronary stenoses compared between both angiograms were set in relation to a number of conventional stenosis parameters at baseline. Regression analysis demonstrated a significant correlation of the changes in minimal diameter with baseline % diameter stenosis (r = 0.30; P < 0.001), minimal diameter (r = -0.28; P < 0.001) and reference diameter of stenoses (r = -0.14; P < 0.001). The changes were not correlated with stenosis length and plaque area. The baseline parameters of 22 preexisting stenoses progressing to occlusions differed from those remaining patent only with regard to the % diameter stenosis (43 +/- 9% vs 39 +/- 11%; P < 0.05). Additional progression of coronary disease became manifest through development of 228 stenoses and 19 occlusions at arterial sites free from definitive stenoses in the baseline angiograms. Thus, progression of atherosclerosis predominantly occurred in mild preexisting coronary stenoses and developed at previously angiographically normal sites. Since the conventional angiographic parameters analysed in this study failed to identify individual arterial sites with an increased risk for progression, definition of new angiographic parameters or application of new techniques seem mandatory to this end.