This project delves into predicting disease using linear regression and cross-validation methods. The objective is to accurately predicts disease, tested through rigorous evaluation metrics and cross-validation for robustness. The methodology encompasses handling missing values, partitioning the dataset, implementing linear regression, and employing various error metrics for assessment. Crucially, cross-validation is utilized to gauge the model's stability and generalizability. The outcome reveals the model's initial promising performance, tempered by insights from cross-validation, emphasizing the importance of generalization in predictive accuracy. This study not only offers valuable insights into heart disease prediction but also establishes a methodological framework applicable to similar medical datasets.