The measurement of collateral flow reserve (CFR; the hyperemic/baseline collateral flow velocity ratio) in patients with chronic total coronary occlusion requires invasive and expensive techniques. Noninvasive transthoracic coronary Doppler echocardiography may be an alternative option. Fifty-one patients with chronic total coronary occlusion were evaluated by transthoracic coronary Doppler echocardiography and venous adenosine infusion to measure CFR in occluded coronary arteries (the left anterior descending artery in 44 patients and the artery supplying the posterior descending artery in 7 patients). CFR data were plotted against 3 angiographic parameters: (1) grade of the epicardial filling of the occluded artery (1=absent, 2=partial, 3=complete), (2) stenosis of the donor artery, and (3) the extent of coronary artery disease (vessels with >or=70% stenosis). Collateral flow was maintained at stress in 34 patients (CFR>or=1, range 1.0 to 2.2) but was withdrawn in 17 patients (CFR<1, range 0.25 to 0.90). CFR increased with the degree of angiographic collateral flow (grade 1: 0.73+/-0.29; grade 2: 1.16+/-0.31; grade 3: 1.34+/-0.49; F=5.31, p=0.008). A multivariate model of CFR prediction showed a direct relation with angiographic collateral grade and the number of diseased vessels and an inverse relation with stenosis of the donor artery. In conclusion, CFR measurement is feasible by transthoracic coronary Doppler echocardiography. One third of the patients with chronic total coronary occlusion had collateral flow withdrawal at stress, which occurs when collateral circulation is poor and when the donor artery is stenotic. CFR correlates with angiographic collateral grade and with the extent of coronary artery disease.