This study evaluated the clinical, intravascular ultrasound (IVUS), and angiographic determinants of the coronary flow reserve (CFR) as measured by guidewire Doppler velocimetry. Using standard methodology, 86 consecutive patients were studied before intervention (n = 73 patients, including the assessment of intermediate stenoses) and/or after intervention (n = 39 patients, including after percutaneous transluminal coronary angioplasty (PTCA) in 27 and post-Palmaz-Schatz stent placement + high-pressure adjunct PTCA in 12). Only 5 patients were studied before intervention, post-PTCA, and poststent. Univariate and multivariate clinical, quantitative coronary angiography (QCA), and IVUS correlates of the CFR were evaluated. There was a linear relation between CFR and IVUS minimum lumen cross-sectional area (CSA): r = 0.771, p <0.0001 for the overall cohort; r = 0.831, p <0.0001 before intervention; r = 0.514, p = 0.0061 post-PTCA; and r = 0.623, p = 0.0306 poststent placement. Overall, an IVUS minimum lumen CSA of > or = 4.0 mm2 had a diagnostic accuracy of 89% in identifying a CFR of > or = 2.0. This diagnostic accuracy increased slightly to 92% when only the preintervention observations were considered. Using multivariate linear regression analysis, the independent determinants of the CFR in the overall cohort of 112 observations were IVUS minimum lumen CSA (p <0.0001), angiographic lesion length (p = 0.0101), and diabetes mellitus (p = 0.0371): r2 = 0.6224. When the subset of preintervention observations were analyzed separately, the independent determinants of the CFR were minimum lumen CSA (p <0.0001) and angiographic lesion length (p = 0.0095); r2 = 0.7176. Thus, the major determinants of the CFR in patients with coronary artery disease are lumen compromise (which is best assessed by the IVUS measurement of the minimum lumen CSA) and lesion length. A minimum lumen CSA > or = 4.0 mm2 has a high diagnostic accuracy in predicting a CFR > or = 2.0, especially before intervention.