The present study examined whether Thoroughbred horses performing strenuous exercise exhibit intrapulmonary arteriovenous shunting which may contribute to the observed arterial hypoxemia. Experiments were carried out on 7 healthy, exercise-trained Thoroughbreds at rest, maximal exercise (galloping at 14 m/s on a 3.5% uphill grade for 120 s), and submaximal exertion (8 m/s on a 3.5% uphill grade for 150 s). Along with blood-gas/hemodynamic parameters, intrapulmonary arteriovenous shunting was studied by injecting 15 µm diameter microspheres, labeled with different stable isotopes, into the right atrium while simultaneous blood samples were being withdrawn at a constant rate from the pulmonary artery and the aorta. Arterial hypoxemia was observed only during maximal exercise. Also, despite significant pulmonary arterial hypertension during submaximal and maximal exertion, 15 µm microspheres did not traverse the pulmonary microcirculation to appear in the aortic blood. Thus, our findings did not support a role for intrapulmonary arteriovenous shunts >15 µm in diameter in the exercise-induced arterial hypoxemia in racehorses. Interestingly, our observation that in going from 30 to 120s of maximal exertion, arterial O₂ tension had remained unchanged despite significant reductions in mixed-venous blood O₂ tension, hemoglobin-O₂ saturation and O₂ content, also discounts the importance of intrapulmonary arteriovenous shunts in causing arterial hypoxemia. This is because, assuming a constant fraction of total pulmonary blood flow bypasses the gas-exchange areas of the equine lungs via intrapulmonary arteriovenous shunts during 30 to 120 s of maximal exertion, the observed significant reductions in mixed-venous blood oxygenation should cause a significant reduction in arterial O₂ tension, which was not the case in our horses. Thus, it is suggested that intrapulmonary arteriovenous shunting probably does not contribute to the exercise-induced arterial hypoxemia in racehorses.