In order to solve problems that the traditional demodulated resonance technology is difficult to determine the parameter of the band-pass filter and greatly influenced by noise,an adaptive demodulated resonance based on EEMD,information entropy and fast kurtogram is presented. Firstly,the original signal is decomposed into several IMFs,and the signal is reconstructed by using the correlation coefficient method to highlight the fault characteristic signal. Then,select the minimum information entropy of IMF to spectral kurtosis analysis and the central frequency and bandwidth of the band-pass filter is determined. Lastly,the filtered reconstructive signal is analyzed by using the Hilbert envelope spectrum analysis. The numerical simulation signal and measured gear fault signal test show that the fault characteristic is highlighted by using the proposed method and the validity is proved.