Studies in animals with normal coronary arteries have shown that coronary flow reserve can be perfected by angiographic measurements of arterial stenosis. Studies in man, however, suggest that even quantitative analysis of coronary angiograms cannot predict the physiologic significance of individual coronary lesions. These studies, however, were carried out in patients with either widespread, diffuse coronary artery disease or by measurement techniques that tend to underestimate maximal coronary flow reserve. To determine the relationship between coronary arterial stenosis and coronary flow reserve (CFR) in patients with discrete limited coronary atherosclerosis, we studied 50 patients with a single discrete coronary stenosis in only one or two vessels. The minimum coronary arterial cross-sectional area (mCSA), percent area stenosis (%AS), and percent diameter stenosis in the left and right anterior oblique projections were determined by the Brown/Dodge method of quantitative coronary angiography. A No. 3F coronary Doppler catheter was placed immediately proximal to the lesion. Measurements of CFR were obtained by intracoronary administration of papaverine in doses sufficient to provide maximal arteriolar vasodilation. In 25 patients, a translesional pressure gradient was obtained with an angioplasty catheter. CFR measured in patients with coronary artery disease was compared with that in 13 patients with normal coronary vessels. In normal patients, CFR averaged 5.0 ± 0.6 (peak/resting velocity ratio; mean ± SEM, range 3.7 to 8.2). In patients with limited coronary artery disease, CFR was closely correlated with %AS (r = .85), mCSA (r = .79), and the translesional pressure gradient (r = .83). Additionally, the most severe percent diameter stenosis in either the left or right anterior oblique view was also highly correlated with CFR (r = .82). Importantly, all arteries with lesions producing less than 70% area stenosis and less than 50% diameter stenosis, or with greater than 2.5 mm2 mCSA had CFR of over 3.5. These results suggest that, in contrast to the poor correlation of percent area and percent diameter stenosis to CFR measured in patients with multivessel coronary artery disease, CFR measured at angiography in patients with discrete, limited coronary artery disease correlated closely with luminal stenosis determined precisely with quantitative coronary angiography. Differences in the extent of diffuse arterial narrowing may account for these discrepancies.