With the development of supercomputer technology, algorithms applied in various fields are also in vogue. Aerobic exercise cardiopulmonary function detection system based on machine learning algorithm is also in a large number of research and development, the prevalence of machine learning algorithm, to medical equipment is intelligent outcome. The intelligence of medical instruments plays an important role in realizing the modernization of clinical diagnosis and rehabilitation health care. Cardiopulmonary function testing system is a set of clinically validated system, which has a reliable and safe guarantee system for patients. This system is bound to bring good news to patients and positive effects to society. This paper studies the cardiopulmonary function detection system of aerobic exercise based on machine learning algorithm, introduces the definition, principle and other related content of machine learning algorithm, and expounds the structure, principle and other related content, definition and concept of cardiopulmonary function detection system of aerobic exercise. Based on machine learning algorithm, the paper uses the effect test method to test the data of cardiopulmonary function detection system of aerobic exercise. Through the test scheme, the final results show that the accuracy and safety of data detection in cardiopulmonary function detection system of aerobic exercise. Portability and efficiency of 86.98%,90.54%,93.86% and 98.35%.