Nowadays health care problem has become an increasingly important topic, which means that more and more people choose to pay for health care insurance. This article is mainly about constructing a model which is used to predict a person’s health care insurance cost as a reason that it is usually hard to find the total amount of cost from people on health care insurance area. The author found a data set about health care insurance cost and used r to analyze it. In the end, 2 linear regression models which seem to fit the data set are built to predict the cost of heal care insurance. After the comparison of these 2 models, a better one is chosen. The model has also been checked to see if it is accurate and proper. The particular significance of this study is that it builds a model which can predict hard-to-collect data from some data which is easy to collect like BMI and age.