Obesity is a worldwide health crisis associated with chronic illnesses such as diabetes and cardiovascular diseases, predominantly fueled by poor dietary habits and inactive lifestyles. Consequently, early detection and effective intervention measures are crucial to alleviate its enduring repercussions. Conventional obesity prediction models predominantly depend on black-box ML methods, hence constraining their interpretability and reliability in healthcare decision-making. XAI4Obesity is an innovative framework that combines Explainable AI (XAI) methodologies with machine learning models to predict obesity risk. This work improves model transparency by employing XGB, RF, SVM, and other classifiers, in conjunction with SHapley Additive Explanations (SHAP), and Local Interpretable Model-Agnostic Explanations (LIME) to improve model transparency. Obesity risk factors—such as weight, height, physical activity, and lifestyle choices—are determined through an empirical dataset. LightGBM outperforms other ML classifiers and attains an accuracy of 97.87%, indicating exceptional performance. The XAI methodology enhances interpretability, cultivating trust in AI-driven healthcare decisions and supporting individualized obesity prevention measures.