A total of 76 splenectomized dogs were entered in a study of the value and effects of isovolemic hemodilution. Of these, seven were not included in the analysis because of technical errors. Of the remaining 69 dogs, 35 were treated with hemodilution; 28 were subjected to a 6-hour period of temporary occlusion of the distal internal carotid artery and the proximal middle cerebral artery, and seven underwent a sham operation only, with arterial manipulation but no occlusion. The other 34 dogs were not subjected to hemodilution; 26 of these underwent temporary arterial occlusion and eight had a sham operation only. In each group the animals were about equally divided into 1) an acute protocol with regional cerebral blood flow measurements by a radioactive microsphere technique and sacrifice at the end of the acute experiment, and 2) a chronic protocol with survival for 1 week to permit daily neurological assessment and final histopathological examination but without blood flow measurements. Isovolemic hemodilution was performed about 1 hour after the arterial occlusion or sham operation and was accomplished by phlebotomy and infusions of low molecular weight dextran to bring the hematocrit to a level of 30% to 32%. This treatment resulted in a very significant reduction in viscosity and fibrinogen levels. The decrease in hematocrit lasted throughout the week in the animals in the chronic protocol. The decrease in viscosity correlated almost linearly with the decrease in hematocrit. There was a slight decrease in systemic arterial pressure with hemodilution but there were no significant changes in central venous pressure or in pulmonary arterial or wedge pressure. There was a slight decrease in cardiac index in both the hemodilution and control groups, which may have been due to the effects of barbiturate anesthesia. There was a slight increase in the measured blood volume in both groups, which was probably artifactual and related to the method of calculation. Intracranial pressure increased significantly with time in all animals subjected to arterial occlusion, but this increase was less severe in the hemodilution group. There was no significant change in intracranial pressure in sham-operated animals, whether hemodiluted or not. The results of cerebral blood flow measurements, assessment of neurological conditions, and measurement of infarct size are given in Part 2 of this report.