In order to more accurately identify the impact of different factors on athletes’ health status and competitive performance, develop personalized training plans and rehabilitation programs and use CART decision tree to carry out research on multisource medical data mining methods for track and field sports. First, use a variety of data sources to collect athletes’ sports data and health information and preprocess the collected data. Second, the association rules of data mining are designed to find the association and correlation between item sets from a large amount of data. On this basis, the CART decision tree is constructed to mine multisource medical data of track and field sports in an all-round and multidimensional manner. The experimental results show that the proposed method can not only accurately identify the data but also fully cover the data in medical data mining, and the application effect has significant advantages.