It is generally belleved that blood flow (CBF) in a severely stenotic coronary artery does not change, or only increases to a maximum limited by the obstruction, in response to increased myocardial oxygen demand in coronary artery disease (CAD). The effects of ventricular pacing on pressure and flow in critically stenotic coronary arteries were studied in six open-chest dogs. A critical stenosis was produced in the circumflex coronary artery with a soft wire snare. CBF in the circumflex (Cx) and left anterior descending (LAD) coronary arteries were measured with electromagnetic flow probes. Aortic pressure, coronary pressure distal to the stenosis, and heart rate (HR) were also measured. In Cx arteries with no stenosis, increasing HR from 59 ± 6 to 153 ± 5 bpm increased CBF from 26 ± 2 to 44 ± 3 ml/min (p < 0.01). When the snare was tightened so that a critical stenosis was produced, increasing HR from 65 ± 3 to 157 ± 4 bpm caused a fall in CBF in the stenotic Cx artery from 23 ± 2 to 15 ± 2 ml/min (p < 0.001). This fall in CBF was associated with a decrease in distal Cx artery pressure (66 ± 3 to 31 ± 2 mm Hg; p < 0.001) and a rise in large artery resistance (2.1 ± 0.4 to 7.7 ± 2.3 units; p < 0.05). In these experiments, CBF in the normal LAD increased from 29 ± 3 to 55 ± 7 ml/min (p < 0.001). Mean aortic pressure did not change significantly (110 ± 3 to 105 ± 4 mm Hg). The fall in Cx flow and distal Cx pressure tended to persist after cessation of pacing. These studies suggest that an increase in myocardial oxygen demand may result in a paradoxical fall in flow in a stenotic coronary artery, and that this reduction in flow may persist after the metabolic stress is relieved.