A method is presented for unbalance fault diagnosis of helicopter rotor by using fuzzy radial basis function （ R BF） neural network. A diagnosis model based on fuzzy RBF neural network is establ ished and a model test is conducted on the rotor. Three types of rotor system faults are considered: imbalanced mass, misadjusted pitchcontrol rod and misadjusted trim tab. Power spect rum is applied to the data p rocessing, and the imbalance fault feature is extracted by using principal comp onent analysis (PCA). The fuzzy RBF neural network diagnosis model is employed t o identify the rotor unbalance faults and compared with the diagnosis model base d on RBF neural network and support vector machine(SVM). The results show that the fu zzy RBF neural network diagnosis model is better than RBFbased model and SVM model in diagnosing the unbalance faults of helicopter rotor.