In recent years, digital twin as an emerging artificial intelligence technology has provided greater potential prospects for precision medicine. Digital twin can build virtual entities by using comprehensive data information of entities, and improve the accuracy of classification and prediction models through the dynamic connection between entities and virtual entities. At present, the application of digital twin in precision medicine not only includes the treatment of specialized diseases, but also includes the health management of entire life cycle of all population. However, most applications are at the stage of technical model design and the validation of single-center data, and greater potential value need to be developed. This paper summarizes the research progress and challenges of digital twin application in precision medicine, so as to provide directions for overcoming the technical bottleneck, expanding the application field, accelerating the implementation of application, and strengthening laws and regulations.