It has previously been demonstrated that cardiopulmonary bypass (CPB) causes prolonged pulmonary vascular hyperreactivity (D.P. Nyhan, J.M. Redmond, A.M. Gillinov, K. Nishiwaki, and P.A. Murray. J. Appl. Physiol. 77: 1584-1590, 1994). This study investigated the effects of CPB on endothelium-dependent (acetylcholine and bradykinin) and endothelium-independent (sodium nitroprusside) pulmonary vasodilation in conscious dogs. Continuous left pulmonary vascular pressure-flow (LP-Q) plots were generated in conscious dogs before CPB and again in the same animals 3-4 days post-CPB. The dose of U-46619 used to acutely preconstrict the pulmonary circulation to similar levels pre- and post-CPB was decreased (0.13 +/- 0.01 vs. 0.10 +/- 0.01 mg.kg-1.min-1, P < 0.01) after CPB. Acetylcholine, bradykinin, and sodium nitroprusside all caused dose-dependent pulmonary vasodilation pre-CPB. The pulmonary vasodilator response to acetylcholine was completely abolished post-CPB. For example, at left pulmonary blood flow of 80 ml.kg-1.min-1 acetylcholine (10 micrograms.kg-1.min-1) resulted in 72 +/- 15% reversal (P < 0.01) of U-46619 preconstriction pre-CPB but caused no change post-CPB. However, the responses to bradykinin and sodium nitroprusside were unchanged post-CPB. The impaired pulmonary vasodilator response to acetylcholine, but not to bradykinin, suggests a selective endothelial defect post-CPB. The normal response to sodium nitroprusside indicates that cGMP-mediated vasodilation is unchanged post-CPB.