This study tested the hypothesis that a reduction in blood viscosity by means of isovolumetric hemodilution will permit an increase in maximal oxygen delivery to myocardium distal to a moderate coronary arterial stenosis. It is known that blood viscosity is a determinent of resistance to blood flow at both the stenotic and the arteriolar levels. Accordingly, a reduction in blood viscosity could exert a favorable influence on maximal myocardial oxygen delivery in the setting of stenosis, provided that the oxygen-carrying capacity of the blood is not compromised excessively. Closed-chest, sedated domestic swine (n = 8) were instrumented with an artificial coronary arterial stenosis that reduced vessel diameter by 64%. Measurements of hemodynamics, regional myocardial blood flow (microspheres), lactate and oxygen metabolism, and whole blood viscosity were made at control and after two successive 10 min intracoronary infusions of adenosine (400 and 800, ug/min) distal to the stenosis. Next, albumin/saline solution was given intravenously to reduce the animal's hematocrit by approximately 50%. Repeat measurements of all experimental variables were then made at a second control and again after two successive 10 min intracoronary infusions of adenosine (400 and 800 gg/min) distal to the stenosis. Myocardial blood flow (ml/min/g) distal to the stenosis increased from 1.52 + 0.21 (mean 1 SD) to 4.10 0.86 in response to adenosine (peak dose) before hemodilution (p < .01) and from 2.07 + 0.59 to 4.08 + 0.93 (p < .01) after hemodilution. Minimum resistance (mm Hg/ml/min/g) distal to the stenosis, however, was approximately 33% lower (p < .05) during infusion of adenosine after hemodilution than it was before hemodilution (endocardium 15.8 6.3 vs 24.5 14.1 and epicardium 9.0 ± 2.3 vs 14.0 ± 8.0). Maximal oxygen delivery (ml/min/lOOg) to myocardium distal to the stenosis failed to improve and in fact was reduced (p < .01 vs before hemodilution) after hemodilution (34.6 ± 9.5 vs 19.9 ± 6.8 to endocardium and 65.5 ± 16.4 vs 38.0 + 10.5 to epicardium). Regional myocardial lactate metabolism, however, did not change vs initial control during the study. Finally, hematocrit was reduced from 32 3% to 17 + 3% (p < .01) and blood viscosity was reduced from 3.4 ± 0.2 to 2.4 ± 0.3 centipoise (p < .01) by hemodilution. The results of the study indicate that reducing blood viscosity by isovolumetric hemodilution may not enhance maximal myocardial oxygen delivery in the setting of a moderate coronary arterial stenosis. However, because minimal endocardial resistance is lowered by a reduction in blood viscosity, it is likely that maximal oxygen delivery could be improved by this intervention if hemodilution were accomplished with a fluid capable of transporting oxygen (e.g., perfluorocarbon emulsion).