Abstract: Physiological causes that influence electrical parameter values of post-harvest apples were analyzed in order to obtain electrical properties of fruits and vegetables and provide new methods for automatically evaluating quality of fruits and vegetables. Based on changes of electrical properties, BP neural network technique was used to distinguish freshness of apples. Results showed the change regularities of relative dielectric constant and the amount of released ethylene from Fuji apples were similar, and both of them were reversed to resistivity. When the amount of released ethylene was at climacteric, relative dielectric constant was maximum and resistivity reached the minimum. Apples were divided into three freshness degrees according to the changes of total soluble solid (TSS) and pH values. The relative dielectric constant and resistivity were selected as the input data of BP neural network, and the average distinguishing rate for apple freshness was 79% under the 2-20-3 network structure. The study provides new evidence for understanding the causes of electrical property changes of post-harvest apples, and offers experimental basis for studying methods to distinguish internal quality of fruits based on electrical properties.