ObjectiveTo investigate the effects of epigallocatechin-3-gallate （EGCG） on learning and memory abilities of amygdala electrical kindling-induced epilepsy in rats and its mechanism. MethodMale SD rats were randomly divided into the normal group， model group， intervention group （model+25 mg·kg^-1 EGCG）， and EGCG group （25 mg·kg^-1 EGCG）. Rats in the EGCG group were only given EGCG intraperitoneal injection， those in the normal group were only given electrode implantation， and those in the other experimental groups were given amygdala electrical kindling stimulation to establish a chronic kindling epilepsy model. EGCG was injected intraperitoneally daily before electrical stimulation. Twenty-four hours after the last electrical stimulation， the escape latency and percentage of target quadrant were recorded by the Morris water maze. Twenty-four hours after the behavioral test， rats in each group were sacrificed by decapitation. The number of hippocampal neurons was observed by Nissl staining. The thickness of postsynaptic density in the hippocampus， synaptic cleft， length of active zone and the curvature of synaptic interface were observed by transmission electron microscopy （TEM）. The expressions of synapse-related proteins synaptotagmin （Syt）， postsynaptic density-95 （PSD-95） and Kalirin-7 in the hippocampus were examined by Western blot. ResultCompared with those in the normal group， the escape latency was significantly prolonged （P<0.05， P<0.01） and the target quadrant ratio was significantly decreased in the model group （P<0.05）. The number of hippocampus neurons decreased significantly （P<0.01）. The synaptic cleft of the hippocampus was widened significantly， and the length of active zone and the thickness of postsynaptic density were significantly decreased （P<0.05， P<0.01）. The expressions of synapse-related proteins Syt， PSD-95 and Kalirin-7 in the hippocampus were significantly decreased （P<0.05，P<0.01）. Compared with those in the model group， the escape latency was significantly shortened and the percentage of target quadrant was significantly increased in the intervention group （P<0.05， P<0，01）. The number of hippocampal neurons significantly increased （P<0.01）. The synaptic cleft of the hippocampus was significantly shortened， and the length of active zone and postsynaptic density were significantly increased （P<0.05， P<0.01）. The expressions of synaptic related proteins Syt， PSD-95 and Kalirin-7 were significantly increased （P<0.05， P<0.01）. ConclusionEGCG can effectively improve cognitive dysfunction after epilepsy. Its protective effect may be achieved by protecting the ultrastructure of hippocampal synapses and regulating the expressions of synapse-related proteins Syt， PSD-95 and Kalirin-7.