We hypothesized that left atrial hypertension results in pulmonary vasoconstriction, which is obscured by the expected passive decrease in pulmonary vascular resistance. The objectives of this study were to demonstrate and quantify the vasoconstrictive changes that occur in the pulmonary circulation during experimental left atrial hypertension, to determine the site of vasoconstriction, and to explore its mechanism. Sheep were instrumented for measurement of pulmonary arterial (Ppa), left atrial (Pla), and systemic arterial pressures (Psa) with a Foley balloon catheter to variably obstruct the mitral valve. Distal pulmonary arterial wedge pressure (Ppaw) was determined by using a 5-Fr Swan-Ganz catheter that was advanced until it wedged with the balloon deflated. Cardiac output (CO) was estimated by thermodilution; pulmonary vascular resistances (PVR) were calculated as mean (Ppa  Pla)/CO = total PVR, (Ppa  Ppaw)/CO = upstream PVR, and (Ppaw  Pla)/CO = downstream PVR. We studied 15 awake sheep at baseline and during increases in Pla of 10 and 20 cmH₂O, with and without inhalation of ~36 parts per million of nitric oxide. Left atrial hypertension resulted in elevation of Ppa. CO decreased only slightly at both levels of Pla elevation. Nitric oxide inhalation caused a significant decrease in PVR, which was greater as Pla increased. This vasodilator effect was most striking in downstream vessels. Experiments with phentolamine, atropine, and ibuprofen failed to reveal the mechanism of the reactive pulmonary vasoconstriction.