To solve the problem of aero-engine multi-performance degradation modeling and reliability evaluation, an evaluation method based on data fusion of active component and stochastic process was proposed. First, multivariate performance degradation data are reduced to low-dimensional key variables through data fusion techniques; then, based on the above variables, the reliability of the product is evaluated using the Gamma process or multivariate Gamma process model; finally, the product life distribution is predicted based on the reliability results at different time. The proposed method improves the efficiency of complex data processing and improves the applicability and practicability of the model.