Abstract: Along with the development of science and technology, the technologies of EMU equipment maintenance were improved. With the accumulation of running time, the probability of equipment failure was also increased. Condition-based maintenance strategy was to collect the equipment running real-time data, compare it with the normal value, find the potential failures and predict the life. Based on the strategy, it was established a new model for key components life prediction. The data was calculated with MATLAB. The results showed that the model could be applied to the confirmation of the running time of EMU.