In order to evaluate the insulation state of the arrester during operation. A a high-voltage experimental system of arrester is set up in the laboratory to simulate the aging of arrester under operating conditions. The test voltage is 5kV ~ 13.6kv, the temperature is 20 ℃ ~ 60 ℃, the relative humidity is 20% ~ 80%, and the salt density is 20g / L ~ 160g / L. In addition, the accelerated thermal aging test of arrester at 200 ℃ for 100h was carried out. During the experiment, the leakage current and its resistance component of ZnO arrester are monitored in real time and taken as the main characteristic quantity. The BP neural network model is established. The leakage current and its resistance component and environmental variables are taken as the input quantity and the insulation state is taken as the output quantity to evaluate the insulation state during the aging process of arrester.The experimental results show that the relative error of the model is less than 1%, and it can effectively remove the interference of environmental factors on the evaluation of the insulation state of the arrester.