A meta-analysis of 13 randomized placebo-controlled trials demonstrated a benefit for dipyridamole therapy, particularly with longer duration of treatment. Although the mechanism of this effect is not well understood, dipyridamole increases endogenous tissue adenosine, which may have a beneficial effect on myocardial perfusion. Therefore, we measured the effects of dipyridamole on myocardial blood flow (MBF) and coronary flow reserve (CFR) by using positron emission tomography and H2O in patients with coronary artery disease.Forty-four patients with angiographically documented coronary artery disease were double-blind randomized to either oral dipyridamole [200 milligrams (mg) twice daily (bd)] or placebo as add-on to conventional antianginal treatment for 24 weeks. MBF was measured at rest and during dobutamine stress at baseline and study completion for the region subtended by the most severe coronary artery stenosis (Isc) and remote myocardium subtended by arteries with minimal or no disease (Rem). CFR was calculated as MBF-peak/MBF-rest.Thirty-five patients completed the study. Isc MBF-rest decreased in patients receiving dipyridamole (0.10 mL/minute/g; P = 0.03) and increased in the placebo group (0.16 mL/minute/g; P = 0.01) during the 24-week study. No significant change in MBF-peak was demonstrated in either group. Consequently, Isc-CFR increased significantly in patients receiving dipyridamole (1.65 +/- 0.47 vs 1.83 +/- 0.67; P < 0.05). By contrast, Isc-CFR decreased significantly in those receiving placebo (1.74 +/- 0.44 versus 1.38 +/- 0.46; P < 0.03). No change was seen in Rem-CFR territories.At the end of treatment, a reduction in baseline MBF but no significant changes in hyperemic MBF were observed in ischemic myocardial territories, and therefore the significance of the observed improvement in CFR remains unclear.