For the damage degree recognition of bearing inner and outer ring under the single failure mode, an evaluation method based on the Lempel-Ziv index and correlation analysis is proposed. The influence of noise on Lempel-Ziv index is reduced by correlation analysis on the basis of retaining frequency components in the signal. Firstly, the autocorrelation analysis of the original signal is used to reduce the noise component of the signal. Secondly, the coding sequence of the original signal is obtained by 0-1 encoding of the noise reduction signal. Finally the sequence after coding is used to calculate the Lempel-Ziv index to get the signal complexity. The effectiveness of the proposed method is verified by simulation and experimental data. Compared with the traditional Lempel-Ziv complexity and the Lempel-Ziv complexity after filtering, the proposed method can identify the bearing fault signal complexity in the noise environment, whic