OBJECTIVE： To study the effect of EGCG-Zn chelated by epigallocatechin gallate （EGCG） and Zn ion on learning， memory and antioxidant abilities in vascular dementia （VD） rats. METHODS： Rats were randomly divided into sham group， model group， zinc gluconate group （positive control， 70 mg/kg）， EGCG group （5 mg/kg）， and EGCG-Zn low-dose， middle-dose， high-dose groups （2， 5， 10 mg/kg）， 10 in each group. The VD models were induced by middle cerebral artery occlusion. The rats were intragastrically given relevant medicines after 7 d of modeling， once a day， for 28 d. Morris water maze were used to evaluate learning and memory abilities of rats. Spectrophotometric method was adopted to detect the superoxide dismutase （SOD）， catalase （CAT）， glutathione peroxidase （GSH-Px） activities and malondialdehyde （MDA） content in brain tissue of rats. RESULTS： Compared with sham group， escape latency was extended in model group， percentage of platform quadrant swimming distance （dP/dT） was reduced （P＜0.01）； SOD， CAT， GSH-Px activities in brain tissue were reduced， and MDA content was increased （P＜0.01）. Compared with model group， escape latency on 4th d of recording was obviously shortened in EGCG group and EGCG-Zn dose groups， dP/dT was obviously increased and dose-depended （P＜0.05 or P＜0.01）； SOD， CAT， GSH-Px activities in brain tissue were increased， MDA content was decreased （P＜0.05 or P＜0.01） and dose-depended. Compared with EGCG group and zinc gluconate group， escape latency was obviously shortened in EGCG-Zn dose groups， dP/dT was increased； SOD， CAT， GSH-Px activities in brain tissue were increased， MDA content was decreased； and the effects in high-dose， medium-dose groups were more obvious （P＜0.05 or P＜0.01）. CONCLUSIONS： EGCG-Zn can improve the learning， memory and antioxidant abilities in VD rats， and the effect is superior to EGCG.