The vasodilation responses to increased blood flow and acetylcholine require endothelial cells. We noticed that dogs with low hematocrit had reduced endothelial cell dependent responses; infusion of whole blood often restored the responses. Therefore, experiments were designed to test the hypothesis that hemodilution attenuates endothelial cell-dependent dilation. An extracorporeal shunt was created from the femoral artery to the jugular vein in pentobarbital-anesthetized dogs. Femoral artery diameter was measured by sonomicrometry. Blood flow was controlled by a screw clamp placed distally on the shunt tubing, and flow was increased from control (10% maximum) to maximum by opening the clamp for 3 mins. Hemodilution was achieved by withdrawal of whole blood and infusion of either 1) saline, 2) physiologic salt solution (PSS, containing 2.7 mM CaCl2), 3) saline with CaCl2 (2.7 mM), or 4) PSS without CaCl2. Endothelial cell-dependent dilations were evaluated after a 50% decrease in hematocrit. Hemodilution with either PSS or saline with CaCl2 did not decrease dilation responses to increased flow or acetylcholine. However, hemodilution with saline or PSS without CaCl2 markedly attenuated endothelium-dependent dilations. Ionized plasma calcium concentration decreased with saline and PSS without CaCl2 hemodilution, but it was maintained with PSS and saline with CaCl2 hemodilution. These data suggest that a 50% decrease in hematocrit does not influence endothelium-dependent dilation if plasma calcium is maintained. Our data support in vitro results that suggest that extracellular calcium is necessary for the release of endothelium-derived relaxing factor. Furthermore, relatively small changes in ionized calcium, within the physiologic range, have large effects on endothelial cell-mediated dilator responses to flow and acetylcholine. However, relatively large changes in hematocrit have no effect on the endothelium-dependent responses.