Abstract: Materials service evaluation encompasses multiple aspects including performance characterization, failure analysis, and remaining life prediction, playing a crucial role in ensuring the safe operation of major engineering equipment and optimizing maintenance strategies. As the service environments of materials in key engineering fields such as aerospace, nuclear energy, and transportation become increasingly complex, traditional materials in-service evaluation methods exhibit certain limitations in terms of real-time capability, accuracy, and intelligence. Digital twin technology offers a novel solution by integrating physical models, data-driven approaches, and real-time monitoring. This technology enables dynamic assessment of material service conditions and performance prediction. Among various failure modes during service, corrosion, fatigue, and fracture are the most prevalent. These modes directly impact the operational safety of critical infrastructure and equipment. This paper first systematically reviews the application progress of digital twin technology in the research of these three typical failure modes. Subsequently, focusing on the challenges faced by digital twins for materials in-service evaluation, it provides an in-depth analysis of the research status of key technologies including multi-source data fusion, multi-scale modeling, real-time data transmission, and service evaluation model construction. Finally, the future development trends of digital twins for materials in-service evaluation are discussed.