OBJECTIVE We evaluated isovolumic hemodilution with hydroxyethyl starch 200/0.5 in a rat model of focal cerebral ischemia. This compound avoids the unfavorable viscosity and erythrocyte aggregation abnormalities of low molecular weight dextran during administration over a period of several days. METHODS Sprague-Dawley rats, anesthetized with 0.5-1% halothane and 70% N2O, were subjected to silicon cylinder (treated and control groups) or sham (sham group) embolization of the cerebral circulation. Thirty minutes after embolization, the treated group (n = 5) was infused with 11 ml/kg of 10% hydroxyethyl starch 200/0.5, and the control (n = 5) and sham (n = 4) groups were infused with saline for 1 hour. In the treated group, 7.1 ml/kg of blood was withdrawn. After 24 hours, the animals were reanesthetized, and cerebral blood flow was determined with [14C]iodoantipyrine. Alternative brain slices were either incubated with 2,3,5-triphenyltetrazolium chloride for infarct volume determination or frozen for ischemic volume and cerebral blood flow determination using autoradiography. RESULTS The hematocrit in the treated group was reduced from (mean +/- SEM) 46 +/- 1% to 35 +/- 2% at 1.5 hours (p < 0.01). Cortical blood flow was within the normal range of 115-185 ml/min/100 g, except for the ischemic cortex in the embolized groups, treated and control. The ischemic and infarct volume of the treated group was reduced by 74% (p < 0.05) and 89% (p < 0.05), respectively, from the control group. The treated and sham ischemic and infarct volumes were not statistically different. CONCLUSIONS These data suggest that hydroxyethyl starch 200/0.5 could be an effective treatment for ischemic stroke when administered early, because it reduces infarct and ischemic volumes from control values to levels indistinguishable from those of the sham group.