Objective : Repeated administration of mannitol in the setting of large hemispheric infarction is a controversial and poorly defined therapeutic intervention. This study was performed to examine the effects of multiple-dose mannitol on a brain edema after large hemispheric infarction. Methods : A middle cerebral artery was occluded with the rat suture model for 6 hours and reperfused in 22 rats. The rats were randomly assigned to either control (n=10) or the mannitol-treated group (n=12) in which intravenous mannitol infusions (0.8 g/kg) were performed six times every four hours. After staining a brain slice with 2,3,5-triphenyltetrazolium chloride, the weight of hemispheres, infarcted (IH) and contralateral (CH), and the IH/CH weight ratio were examined, and then hemispheric accumulation of mannitol was photometrically evaluated based on formation of NADH catalyzed by mannitol dehydrogenase. Results : Mannitol administration produced changes in body weight of $-7.6{pm}1.1%$, increased plasma osmolality to $312{pm}8;mOsm/L$. It remarkably increased weight of IH ($0.77{pm}0.06;gm$ versus $0.68{pm}0.03;gm$ : p<0.01) and the IH/CH weight ratio ($1.23{pm}0.07$ versus $1.12{pm}0.05$ : p<0.01). The photometric absorption at 340 nm of the cerebral tissue in the mannitol-treated group was increased to $0.375{pm}0.071$ and $0.239{pm}0.051$ in the IH and CH, respectively from $0.167{pm}0.082$ and $0.162{pm}0.091$ in the IH and CH of the control group (p<0.01). Conclusion : Multiple-dose mannitol is likely to aggravate cerebral edema due to parenchymal accumulation of mannitol in the infarcted brain tissue.