Human amniotic epithelial cells (hAECs), having the characteristics of both embryonic and pluripotent stem cells, have the potential to differentiate into various cells. A good deal of research has explored the clinical therapeutic potential of hAECs; rat amniotic epithelial cells have been reported to ameliorate functional deficits after stroke in rats, likely due to neuronal differentiation and cytokine secretion by these cells. We isolated hAECs and transfected them with glial cell line-derived neurotrophic factor (GDNF) or enhanced green fluorescent protein (EGFP) gene using lentiviral vectors. These cells were then transplanted into the brains of rats subjected to a transient middle cerebral artery occlusion. The hAECs survived and migrated to the ischemic area of rats, and some of the transplanted hAECs expressed the neuronal marker MAP2 and the neuronal progenitor marker Nestin, together with the astrocyte marker glial fibrillary acidic protein, and hAEC-EGFP can significantly ameliorate behavioral dysfunction and reduce infarct volume of ischemic rats. By transfecting the cells with lentiviral vectors, GDNF can be stably overexpressed in hAECs, and hAEC-GDNF can more rapidly rescue the deficits of rats after middle cerebral artery occlusion compared with hAEC-EGFP-treated rats. Moreover, the nontransduced cells also had effects comparable to the GDNF-transduced cells on caspase-3 and lesion volume. Because hAECs are in unlimited supply, and their use is not encumbered by ethical arguments, hAECs have a great advantage for stem cell therapy. This model holds tremendous potential for development into wide use in cell-mediated gene therapy in the future.