The extent of cerebral injury and edema was determined in isoflurane-anesthetized rats (n = 32) after 180 min of middle cerebral artery occlusion (MCAO) and 120 min of reperfusion. One of the following was employed during the occlusion period only: 1) control, mean arterial pressure [MAP = 131 +/- 7 (SD) mmHg] and hematocrit (43 +/- 2%) were not manipulated; 2) hemodilution, the hematocrit was reduced to 30% with 5% albumin (MAP = 104 +/- 19 mmHg); 3) hemodilution-normotension, hemodilution was established, and MAP was maintained at 131 +/- 9 mmHg with phenylephrine; 4) hemodilution-hypertension, hemodilution was established, and MAP increased to 161 +/- 2 mmHg with phenylephrine. Brain injury was determined with 2,3,5-triphenyltetrazolium chloride, and cerebral edema was assessed by microgravimetry. Brain injury and cerebral edema were less in both phenylephrine groups, compared with the control and hemodilution groups (P less than 0.05). These results are consistent with the premise that if normotension is maintained, hemodilution reduces ischemic brain injury and edema. They also indicate that the addition of phenylephrine-induced hypertension to hemodilution therapy results in a further reduction of ischemic injury without exacerbating cerebral edema.