To determine whether or not angiographic coronary morphology can predict future development/progression of narrowing, we reviewed coronary angiograms (CAGs) from 29 patients who underwent CAG studies twice but who had no myocardial revascularization during the period of the studies. The mean age of the patients was 52.9 +/- 8.5 years, and the mean interval between the studies was 25.4 +/- 22.6 months. Mean luminal diameter stenosis of 77 lesions that reduced the diameter by 50% or more on either CAG, but were not totally occluded on the initial CAG, increased from 62% to 79% (p less than 0.01). Progression of stenosis developed in 35 lesions (45%); the progression in 13 lesions of the 35 (37%) was on sites associated with no stenosis or mild stenosis on the initial CAG. No relation was found between the development/progression of stenosis and either its initial severity or elapsed time; however, the prevalence of the development/progression of stenosis was more frequent on sites with possible or probable ruptured atheromatous plaque on the initial CAG than on sites without such plaque (p less than 0.05). The evidence suggests that there is no apparent relation between the future development/progression of coronary narrowing and either the severity of stenosis on the initial CAG or elapsed time but that some of the development/progression can be predicted on the basis of certain specific coronary morphology.