Coronary artery bypass surgery is used widely for treating myocardial ischemia. However, blood flow and flow reserve of normally perfused myocardium subtended by bypass grafts have not been evaluated late after surgery. Also, it is unknown whether pharmacologic vasodilation evokes comparable myocardial flow responses in arterial and venous conduits. Myocardial blood flow was quantified at rest and during dipyridamole hyperemia using N-13 ammonia and positron emission tomography (PET) in 15 patients 9 +/- 3 years after bypass surgery and in 10 healthy volunteers. Blood flow was analyzed in 26 territories subtended by bypass grafts with normal wall motion and normal perfusion. Myocardial blood flow at rest did not differ between patients and controls (0.65 +/- 0.14 vs 0.68 +/- 0.16 ml/ g/min) and was similar in normal myocardium subtended by saphenous vein (n = 16) and internal mammary artery grafts (n = 10; 0.64 +/- 0.13 vs 0.66 +/- 0.15 ml/g/min). However, the hyperemic response in normal myocardium supplied by bypass grafts was less than that in controls (1.61 +/- 0.33 vs 2.04 +/- 0.30 ml/g/min, p <0.005). No differences between territories supplied by venous and arterial conduits were observed (1.61 +/- 0.35 vs 1.63 +/- 0.32 ml/g/min). Normal myocardium subtended by bypass grafts exhibited a lower flow reserve than that in controls (2.54 +/- 0.51 vs 3.16 +/- 0.85, p <0.02). Myocardial flow reserve was almost identical in regions supplied by venous and arterial grafts (2.55 +/- 0.48 vs 2.52 +/- 0.58). The similar reduction in vasodilatory capacity together with the normal PET polar map findings during dipyridamole argue against flow limiting stenoses in both venous and arterial bypass conduits late after revascularization. Rather, nonobstructive proliferative fibrointimal changes of the bypass conduits or atherosclerosis of the native resistance vessels might account for this finding.