The method of reliability modeling and analysis based on performance degradation data provides an important way for the reliability research of long life and high reliability products in modern industry, but the degradation data cannot be obtained adequately during a destructive inspection. To solve this problem, this paper proposes to use both degradation data and marker data to make reliability inference, which can improve the precision of reliability inference, and make residual lifetime distribution prediction available through inspection of marker during working, contributing to the decision making in maintenance, replacement and spare parts. The performance degradation and marker are modeled with a bivariate Wiener process and, the methods for parameter estimation and residual lifetime distribution prediction are provided. Finally, a simulated example is presented to show the effectiveness of the proposed method.