Research of the reliability evaluation and life prediction technology is of great significance to preventing faults and supporting predictive maintenance during rolling bearing running. The traditional reliability analysis methods with a static model based on large numbers of failure date and empirical knowledge are unable to track the degradation process of rolling bearing and accurate reliability assessment and life prediction. In this thesis, the dynamic reliability analysis model is established based on the principal component analysis (PCA) and phase space reconstruction. The real-time monitoring parameters are fused through PCA, and then the predictive value of life is obtained by comparing the current degradation process with the historical degradation process. With the continuous accumulation of observed samples, the life of rolling bearing can be updated. The experimental result shows that the dynamic life prediction model proposed in this paper can predict the life of the rolling bearing in real time.