Teenager’s obesity is public health issue that needs attention because of the increasing risk of other diseases such as diabetes. This study purposes to build a system that can predict the risk of teenager’s obesity. The data is in the form of a secondary dataset that has 17 features that include eating habits, physical activity, and others. Random Forest is used because of its ability to handle high-dimensional data and produce accurate classifications. This system is console-based, where users can add their lifestyle data and get results in the form of obesity predictions with low, medium, and high levels. The results of the model evaluation show very good results, which 90%  accurate and high consistency, recall, and f1-score values. This method shows stable and competitive performance compared to other algorithms such as Decision Tree and KNN. So the results are expected to be used as a learning and prevention tool to detect the risk of obesity in adolescents early on.