Purpose: To examine which of stress myocardial flow (MBF) and coronary flow reserve (CFR) determined by (15)-water-PET (PET) correspond most closely with diameter stenosis assessed by quantitative coronary angiography (QCA). Methods: Twenty-three patients with a C40% QCA stenosis underwent baseline and adenosine stress PET. Baseline MBF measures were corrected for myocardial workload and stress MBF and CFR calculated in 17 standard AHA myocardial segments and reassigned to respective feeding vessels. If multiple stenoses, only the most severe stenosis was considered. Pearson’s correlation coefficients were calculated and compared using Fisher r-to-z transformation. Results: A total of 35 QCA stenoses were identified with a mean diameter stenosis of 75% ± 18%, range 45-100%. Stress MBF and CFR in corresponding vascular territories were: 1.84 ± 0.78 and 1.86 ± 1.05 mL/minute/g, respectively. MBF was significantly correlated to QCA stenosis severity: r = -0.46, P = .006, whereas CFR was not, r = -0.28, P = .097. These two r values did not differ significantly, P = .42. In contrast, stress MBF and CFR were significantly correlated (C): r = 0.86, P\.001. Conclusions: The correlation between anatomic and functional stenosis was poor, suggesting that the hemodynamic consequences of angiographically proven coronary stenoses should be examined by functional imaging like PET.