With the advantage of strong self-adaptation, Elman neural network is applied to the fault diagnosis of high voltage electric power metering system.Firstly we extract information of the 8 operation parameters of metering system as input values of the neural network, and use 4-bit binary number to represent the normal circumstance and 9 main types of failures, then we collect samples to train Elman network and select 4 types of failures to test the network, The results show that the Elman network model can effectively diagnose the fault of high voltage electric power metering system.