Objective To further verify the practicality of the experimental animal model induced by mice cerebral ischemia-reperfusion injuryand to investigate the optimal time point of mannitol intervention in cerebral ischemia-reperfusion injury. 　Methods　Thirty-six mice were randomly divided into six groups（n=6）： sham operation group （group A）， control group（ischemia-reperfusion injury group， group B）， and mannitol intervention group at 0, 30 min， 1 h and 1.5 h after reperfusion （group C， D， E， F）. Brains were removed twenty-four hours after reperfusion. Morphological changes， cells survival and cellular degeneration in CA1 area of hippocampus were studied in each group.　Results　Ingroup C， most hippocampal pyramidal cells remained normal morphologically， intact structurally and no intercellular edema . The amount of viable cells was （23.01±1.23） per 100 μm in group A and no difference compared with group A（24.00±0.75）. The frequency of cellular degeneration was 8.26%， significantly less than that of group B（30.59%，P<0.05）. And no difference compared with group A（7.45%）. The apoptosis index（AI） was 6.50%， significantly less than that of group B（28.00%，P<0.05）. No difference of compared with that of group A（5.00%）. Mannitol intervention beginning over 30 minutes after reperfusion failed to alleviate neuron injury.　Conclusions　Inverted osmotic pressure gradient by aggressive mannitol intervention can effectively prevent intracellular edema and avoid the subsequent irreversible cellular injury after the establishment of blood pressure gradient at the early stage of cerebral ischemia-reperfusion. The results suggest that during cardiac resuscitation，mannitol dehydration treatment should be initiated simultaneously at the recovery of autonomous heart rate.