Data mining has been widely used in the study of cardiovascular disease prognosis. For stroke prognosis, the focus was mainly on the prediction of intervention effectiveness. In contrast, the focus was primarily on predicting natural prognostic and intervention safety for other cardiovascular diseases. In addition, compared with traditional statistical methods, machine learning, especially deep learning based on neural networks has much better performance in predicting the prognosis of cardiovascular diseases, which is worthy of further promotion and application. Therefore, this study systematically reviewed the recent application progress of data mining in cardiovascular disease prognosis, summarized the shortcomings of current studies, and put forward future directions.