To obtain a mathematical model capable of predicting the 50% discharge voltage of air-gap under high-altitude conditions, we used the test results of air-gap in high altitude areas to establish the 50% discharge voltage BP neural network model of air-gap under high-altitude conditions. We used the model to forecast the 50% discharge voltage of air-gap under high-altitude conditions. The result shows that the maximum error between forecast voltage and test voltage is 1.42%, which certificates the possibility of using the neural network to build the multidimensional and nonlinear relationship between the environmental factors and discharge voltage. At the same time, we can simulate and analyze the effect of environmental factors on discharge voltage of air-gap with the help of model that we established, analysis showed that there was a positive correlation between the environmental factors, such as temperature, humidity and atmosphere pressure, and discharge voltage of air-gap.