To improve the helicopter safety better, the fault diagnosis of swash-plate bearing based on BP neural network and RBF neural network is investigated. Bearing seed fault tests are carried out, vibration data of bearing tests are obtained and its characteristic signals are collected. Some characteristic signals of vibration data are used as the inputs of neural networks to perform the fault diagnosis, and accurate fault diagnosis rates are obtained. The highest diagnosis rates of different kinds of bearing faults are high than 89%.