This paper presents a tool wear monitoring method based on feature fusion via neural networks. The basic principle of this method，feature association，feature extraction，data normalization and neural networks construction of the method are discussed. The experimintal results show that the proposed method for tool wear monitoring is reliable and effective.