Neuroprotective effects of estrogens are demonstrated consistently in the cerebral cortex, but not in subcortical areas. In the present study, transient middle cerebral artery occlusions (MCAO) were induced for various duration, and protective effects of estrogen treatment on the cerebral cortex and subcortex were evaluated. MCAO was induced for 30, 40 or 60 min in ovariectomized rats. Animals were treated with 17beta-estradiol (E2) or vehicle (OVX) 2 h before MCAO and sacrificed 24 h after the indicated duration of MCAO. Ischemic lesion was evaluated by 2,3,5-triphenyltetrazolium chloride staining, hematoxylin and eosin staining, and terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) staining. E2 treatment reduced the magnitude and delayed the appearance of the total ischemic lesion area and largely prevented TUNEL staining in the cortex. In the subcortex, E2 treatment prevented the ischemic lesion in the 30-min group, reduced lesion area in the 40-min group, but had no effect on ischemic lesion area in the 60-min group. E2 treatment significantly decreased apoptotic cell number in the subcortical area at 30 and 40 min, but not at 60 min of MCAO. This study demonstrated that estrogen treatment can protect the cerebral subcortex in a severity-dependent manner, suggesting that the lack of protective effects of estrogen treatment in the subcortex is not due to the lack of estrogen receptors. Further, this study indicates that estrogens could be used as a neuroprotectant to prolong the therapeutic window of thrombolysis and prolong the time of cerebral circulation intervention for neurosurgical procedure.