To explore the effect of low load strengthening on the fatigue life of parts under random load, the automobile stabilizer bar is taken as an example, and the collected random load signal is extrapolated to the whole life cycle by peak over threshold extrapolation method. Based on the extrapolated load signal, the distribution characteristics of its mean and amplitude are analyzed, and the two-dimensional joint probability distribution is established. According to the characteristic that the stabilizer bar is continuously subjected to complex random loads, the influence of mean and amplitude is considered to compile the two-dimensional load spectrum, and the low load strengthening theory is introduced. The life prediction based on linear fatigue cumulative damage theory（Miner rule） and low load strengthening theory is compared with the test results to verify the effectiveness and accuracy of life prediction based on low load strengthening under random load.