In order to effectively evaluate the bearing degradation trend and improve the intelligence of equipment health management, a bearing remaining useful life prediction method based on BAS-BP model was proposed. The time and frequency domain characteristics of bearing life cycle vibration signals were extracted, and the 18 dimensional degradation characteristics were constructed. In order to improve the prediction accuracy of neural network, the initial weight and threshold were optimized by beetle antennae search (BAS) algorithm, and the BAS-BP prediction model was established. The effectiveness of the model was verified on the public data set. The experimental results show that the BAS-BP model can effectively predict the remaining life of the bearing and has high accuracy.