The work aims to analyze the vibration characteristics of the express transport packaging vehicles in south China to provide theoretical guidance for the research of random vibration test. Vibration signal of one terminal delivery is measured for Hilbert Huang Transform. The instinct mode functions decomposed by empirical mode decomposition method are analyzed in aspects of time domain, power spectral density and Hilbert spectral. Empirical Mode Decomposition decomposes random vibration signals of transport packaging into several IMF components in order of frequency. It is very likely that the first several IMFs represent the high-frequency instantaneous events, while the second several IMFs represent the low-frequency vibration signals of the original signal. Thus, the high frequency and low frequency vibration signals can be separated effectively The results show the applicability of HHT for random vibration on express logistics packaging. The impulsive shock events with big amplitude and short duration, which buried in vibration, can be extracted.