The hypothesis was tested that regional myocardial contractile dysfunction can detect subtle regional coronary blood flow maldistribution induced by exercise. In seven dogs, left ventricular pressure (micromanometer), regional systolic wall thickening (WTh, sonomicrometry), and myocardial blood flow (MBF, microspheres) were measured when mild degrees of coronary artery stenosis were produced during treadmill exercise. During exercise without coronary stenosis, WTh increased by 21 ± 12% (SD), and transmural MBF increased uniformly. In each dog, two levels of coronary stenosis were produced during exercise by adjusting the coronary hydraulic cuff: (1) St-Ex I, where WTh during exercise failed to increase significantly (average change 0 ± 7%), and (2) St-Ex II, where WTh during exercise decreased moderately from the resting control value (average −20 ± 8%). In the potentially ischemic zone coronary hyperemia occurred with each run: resting subendocardial MBF was 1.09 0.30 mg/g/min, and it was 3.04 ± 0.83 during control exercise, 2.48 ± 0.75 during St-Ex I, and 1.55 0.59 ml/g/min during St-Ex II (p < .01 compared with control exercise and control area). The subendocardial-subepicardial blood flow ratio fell from 1.32 ± 0.27 during control exercise to 1.07 ± 0.20 (p < .05) during St-Ex I, and to 0.64 ± 0.15 (p < .01) with St-Ex II. Changes in the subendocardial electrogram and reactive hyperemia occurred more consistently during St-Ex II than St-Ex I. Thus, failure of regional function to increase during exercise detected slight maldistribution of regional MBF, whereas reduction of regional function during exercise of 10% or more below the resting value was a reliable marker of a regional flow defect and was always associated with other evidence of ischemia. Therefore, regional dysfunction during exercise can detect subcritical but functionally significant coronary stenosis, which may allow regional wall motion to be used for detecting coronary artery disease at a relatively early stage.