Considering the uncertainty characteristics of the aircraft air refrigerator′s information acquired from project, a lifetime trend analysis method is proposed based on spectrum kurtosis (SK) and least squares support vector machine (LSSVM). First, the life parameter is decomposed into several intrinsic mode functions (IMFs) by empirical mode decomposition (EMD), taking the IMFs for Fourier transform respectively to get the spectrums and calculate the SK of each spectral line. Then through the SK, the important IMFs can be found out those effect on the air refrigerator failure, to signal synthesis. Finally, the energy features of the synthetic signal are extracted to establish the LSSVM trend analysis model to predict the air refrigerator′s life trend. Experimental results show that the proposed method of aircraft air refrigerator life trend analysis based on SK and LSSVM can effectively realize the air refrigerator life trend analysis, with strong potential for engineering applications.