Hypertension is a global health issue that contributes significantly to cardiovascular disease and early death. This study has classified patients as low risk (0) and high risk (1) based on their risk of developing hypertension. The relevant risk label for hypertension we found is 0 for low risk (2581) and 1 for high risk (1170). We applied machine learning techniques, including Stacking Ensemble, Extreme Gradient Boosting (XGBoost), Random Forest Classifier, and Gradient Boosting Classifier, to classify the cases under investigation and identify the most significant features related to hypertension. The models have been evaluated using confusion metrics, such as the precision, recall, F1 score, area under the receiver operator characteristic, area under classification accuracy, and others. In the prediction of hypertension, the stacking ensemble methods performed the best, with an accuracy rate of 0.9161, precision of 0.9026, recall of 0.9069, and F1-score of 0.9047. We think that our study method provides the most well-known answer for predicting control of hypertension, and that, using patient information that is available in the electronic health record, machine learning could increase the accuracy of hypotension control predictions. Our approach may act as the foundation for further research to evaluate the model's utility and increase its accuracy.