Health insurance is a type of insurance in which participants pay regular premiums to ensure coverage for medical expenses with the purpose is to support financial protection to the insured in the event that they need medical services, including treatment costs, medications, and other healthcare expenses. This study has leveraged machine learning to predict healthcare insurance costs. We employ six regression methods to compare and assess the effectiveness of the prediction: Linear Regression, Ridge Regression, Lasso Regression, ElasticNet, Decision Tree, and Random Forest. In addition, we evaluate each factor which can affect the cost. The results indicate that Random Forest performs the best among the models tested.