There is huge clinical value in real world data from traditional Chinese medicine (TCM), but the real world study of TCM faces many challenges, because of its diverse data types, different standards, and serious data island phenomenon. Data governance is the key to transforming real world data into real world evidence. As the last step of data governance, data transformation has not been standardized. The key technologies and methods of data transformation, including data classification, natural language processing, standardization, data system construction, and derivative variables, will be discussed in this article based on the characteristics of real world data of TCM and the current development status of data transformation technology. At the same time, the suggestions of safety control and quality control of data transformation are put forward, and the data transformation system of real world study of TCM is preliminarily constructed, combined with the characteristics of TCM data. It is hoped that this paper can provide references for future real world studies of TCM.