With the continuous progress of research methodology in the real world and the growing maturity of artificial intelligence technology， a method for conducting “quantitative” research to guide clinical practice based on traditional Chinese medicine （TCM） diagnosis and treatment data was gradually developed. However， there is still a need for further improvements in the overall design of studies and the transformation of findings into clinical practice. Based on this， we put forward a comprehensive overall design concept and application approach for real-world study and artificial intelligence research based on clinical diagnosis and treatment data of TCM. This approach consists of five steps： Constructing a research-based database with a large sample size and high data quality； Mining and classification of core prescriptions； Conducting cohort studies to evaluate the effectiveness of core prescriptions； Utilizing case-control studies to clarify the dominant population； Establishing predictive models to achieve precision medicine. Additionally， it is imperative for researchers to establish a standardized system for collecting TCM variables and processing data， optimize the determination and measurement methods of confounding factors， further improve and promote methodologies， and strengthen the training of interdisciplinary talents. By following this research method， we anticipate that the clinical translation of research findings will be facilitated， leading to advancements in TCM precision medicine. Real-world study and artificial intelligence research share similar research foundations， and clinical applications complement each other. In the future， the two will merge together.