Accelerated graft atherosclerosis is a major cause of death after cardiac transplantation. Although its detection currently requires surveillance angiography, loss of vasodilator responsivity may precede obstructive lesions and be detectable by noninvasive assessment of myocardial perfusion. Thirty-five allograft recipients were studied an average of 31 ± 19 (mean ± SD) months after transplantation. All were free from angiographically definable macrovascular obstructive coronary artery lesions. Nutritive myocardial perfusion at rest, estimated in absolute terms by positron emission tomography with oxygen-15 water averaged 1.63 ± 0.51 ml/g/min in patients and was greater than that in 26 healthy volunteers (1.17 ± 0.33 ml/g/min, p < 0.001). The increase correlated with increased cardiac work at rest in transplant recipients with arterial hypertension and tachycardia. Peak myocardial perfusion induced by intravenous administration of dipyridamole was normal in the transplant recipients (3.49 ± 1.70 ml/g/min compared with 3.60 ± 1.41 ml/g/min in volunteers). Because of the high flow at rest, myocardial perfusion reserve (the ratio of hyperemic flow to flow at rest) was diminished (2.3 ± 1.2 compared with 3.3 ± 1.5 in volunteers, p < 0.005). These results indicate that the responsivity to vasodilator stimulation is well preserved in transplant recipients devoid of macroscopic coronary arterial lesions obviating detection of early vascular dysfunction in individual subjects. Positron emission tomography may be useful, however, in quantifying the magnitude of the increase in flow at rest secondary to increased cardiac work - a potentially remedial cause of accelerated coronary vascular disease induced by high shear force activation of platelets in the coronary bed, and in detecting impaired perfusion once macrovascular vascular disease is extant.