We have previously shown that tamoxifen can reduce infarct sizes measured by 2,3,5,-triphenyltetrazolium chloride (TTC) staining at 72 h after 2 h of reversible middle cerebral artery occlusion (rMCAo) in rats. In this study, we tested whether improvement is found in both behavioral measures of protection and by histological measures of infarcted tissue at 7 and 14 days after 2 h rMCAo. Tamoxifen (10 mg/kg) was given once by intravenous injection 1 h after reperfusion, i.e. 3 h after initiation of rMCAo. Neurobehavioral deficits were evaluated daily for 1 week or 2 weeks followed by infarct volumes measurements by hematoxylin–eosin (HE) staining. Tamoxifen-treated rats had significantly improved neurobehavioral deficit scores when evaluated daily throughout the 1 week or 2 week periods and showed significantly reduced median infarct volumes measured after 1 week and 2 weeks. Median infarct values were 149 mm³ (interquartile range, IQR: 92 to 258) and 124 mm³ (IQR: 69 to 174) for the 1 and 2 week vehicle groups, respectively, compared with 5 mm³ (IQR: 3 to 16) and 4 mm³ (IQR: 0 to 48) for the comparable treated groups (both P < 0.05, Mann–Whitney test), giving a reduction of more than 90% in both cases. Thus, a single administration of tamoxifen given 3 h after initiation of rMCAo is extremely effective in producing long-term neuroprotection as assessed by neurobehavioral measures and histopathology in experimental stroke in rats. If these results are extrapolatable to human stroke, these data indicate that tamoxifen may be a useful neuroprotectant.