We tested the hypothesis that a functional abnormality may contribute to impaired myocardial perfusion during tachycardia in the chronically pressure-overloaded hypertrophied left ventricle. Left ventricular hypertrophy (LVH) was produced by banding the ascending aorta of seven young dogs, 5-6 weeks of age, and studies were carried out after the animals reached adulthood when the mean left ventricular:body weight ratio was 91% greater than in seven control dogs. Myocardial blood flow was measured with radioactive microspheres during ventricular pacing at 100, 200 and 250 beats/min in the presence of normal coronary vasomotor activity and during maximum coronary vasodilation (adenosine, iv). At a heart rate of 100 beats/min, there was no difference in myocardial blood flow between control and LVH dogs either during normal conditions or during adenosine infusion. However, since mean coronary perfusion pressure was higher in the dogs with aortic banding, minimum coronary resistance was greater in the dogs with LVH [22.7 ± 2.5 (mean ± SE)] than in the control dogs (14.5 ± 1.4 mm Hg/min per g per ml; P<0.05). In the presence of intact coronary vasomotor tone, pacing at 250 beats/min resulted in transmural redistribution of perfusion away from the subendocardium in the dogs with LVH [subendocardial:subepicardial blood flow ratio (endo:epi) = 0.68 ± 0.09], but not in the control dogs (endo:epi = 1.05 ± 0.03; P<0.02). This relative reduction of subendocardial flow in the dogs with LVH resulted from extravascular factors since blood flow to the inner half of the left ventricular wall failed to increase in response to adenosine infusion during pacing at 250 beats/min. Thus, both an anatomic abnormality of coronary perfusion, marked by impaired minimum coronary resistance at a heart rate of 100 beats/min, and a functional abnormality of perfusion resulting in redistributioh of blood flow away from the subendocardium at a heart rate of 250 beats/min were demonstrated in this model of LVH.