Obesity is a major global health concern that increases the risk of chronic diseases such as diabetes, cardiovascular disease, and cancer while imposing substantial socioeconomic burdens. Traditional obesity diagnosis methods often suffer from limited accuracy and poor interpretability. This study proposes an interpretable machine learning framework for obesity diagnosis using the Kaggle obesity dataset, which includes features related to Body Mass Index (BMI), dietary habits, and lifestyle factors. Several machine learning algorithms, including Random Forest (RF), Gradient Boosting (GB), Support Vector Classifier (SVC), and Extreme Gradient Boosting (XGB), were evaluated, with XGB achieving the highest accuracy of 98%. To enhance transparency and trustworthiness, the Explainable Artificial Intelligence (XAI) techniques SHapley Additive exPlanations (SHAP) and Local Interpretable Model-agnostic Explanations (LIME) were employed. SHAP provided global explanations by identifying BMI, dietary behaviors, and physical activity as the most influential predictors, whereas LIME generated local explanations for individual predictions. Furthermore, an interactive user interface developed using Python Tkinter enables users to input lifestyle data, obtain obesity predictions, and understand the contribution of different features to the model’s decision. The proposed framework combines high predictive performance with interpretability, making it a practical tool for personalized obesity assessment and management.