The clinical and angiographic correlates and the prognostic significance of the "coronary extent score" in a consecutive series of 313 patients who were catheterized twice were studied. The extent score was defined as the number of 5 to 75% stenosed segments in a 15-segment coding system. The extent score was higher in subgroups of patients with new onset angina at the time of the first angiogram (4.3 +/- 2.4 vs 3.3 +/- 1.9, p less than 0.01), unstable angina at the time of the second angiogram (4.0 +/- 2.0 vs 3.3 +/- 1.9, p less than 0.05) or multifocal progression from the first to the second angiogram (4.0 +/- 2.1 vs 3.3 +/- 1.9, p less than 0.01), suggesting that it is an index of active coronary artery disease. The extent score did not correlate with the number of diseased vessels (r = 0.03), the ejection fraction (r = 0.03), the Friesinger score (r = 0.04) and the Gensini score (r = -0.07) (difference not significant for each). Cox's model was fit to the survival data recorded on a prospective basis after the second angiogram. Independent predictors of survival were ejection fraction (p less than 0.001), extent score (p = 0.001), number of diseased vessels (p = 0.01) and percent of left main luminal stenosis (p less than 0.05). The extent score was also an independent predictor of myocardial infarction and unstable angina. Thus, the extent score, an index of active progressive disease, is an independent predictor of mortality and cardiac events in patients with coronary artery disease.