Heart disease could be a leading reason behind premature death within the world.Predicting the end result of malady is that the difficult task.Machine Learning is concerned to mechani- cally infer diagnostic rules and facilitate specialists to create identification method additional reliable.Several data processing techniques are utilized by researchers to assist health care pro- fessionals to predict the heart disease.Random forest is most accurate learning algorithmic rule,suitable for medical applications. .In our project ,we propose a machine learning prediction model that uses random forest algorithmic rule as feature choice measures to predict cardiopa- thy. The experimental results have shown that our approach improve accuracy compared to different approaches.