The in-service life of turbine blades directly affects the on-wing lifetime and operating cost of aircraft engines. It would be essential to accurately evaluate the remaining useful life of turbine blades for safe engine operation and reasonable maintenance decision-making. In this paper, a machine learning-based mechanism with multiple information fusion is proposed to predict the remaining useful life of high-pressure turbine blades. The developed method takes account of the in-service operating factors such as the high-pressure rotor speed and exhaust gas temperature, as well as the engine operating environments and performance degradation. The effectiveness of this method is demonstrated on simulated test cases generated by an integrated blade creep-life assessment model, which comprises engine performance, blade stress, thermal, and creep life estimation models. The results show that the proposed method provides a prospective result for in-service life evaluation of turbine blades and is of significance to evaluating the engine on-wing lifetime and making a reasonable maintenance plan.