Chronic Kidney Disease (CKD) is a global health issue with an increasing prevalence that poses a significant economic burden on healthcare systems. Early detection of CKD is crucial to provide proper treatment before the disease progresses to end-stage renal failure. With technological advancements, machine learning methods have been widely utilized to support medical diagnosis with greater speed and accuracy. This study aims to compare the performance of two popular classification algorithms, Decision Tree C4.5 and Naïve Bayes, in predicting CKD using a public dataset from the UCI Machine Learning Repository consisting of 400 patient records with 24 clinical attributes. The research process involved systematic preprocessing steps, including handling missing values, transforming categorical data into numerical form, and selecting relevant attributes. Model evaluation was conducted using 10-Fold Cross Validation with performance metrics such as accuracy, precision, recall, Area Under the Curve (AUC), and statistical T-Test. The results show that Decision Tree C4.5 achieved an accuracy of 93.00%, precision of 84.27%, recall of 100%, and an AUC of 0.944, while Naïve Bayes obtained an accuracy of 93.50%, precision of 85.23%, recall of 100%, and an AUC of 0.948. Although the performance differences between both algorithms are relatively small and statistically insignificant, Naïve Bayes demonstrated slightly better results in terms of accuracy and AUC, while Decision Tree C4.5 offers advantages in interpretability through its classification rules. In conclusion, both algorithms are effective for early CKD diagnosis, and the choice may depend on practical needs, whether emphasizing interpretability or computational efficiency. This study contributes to the development of more accurate and efficient clinical decision support systems for improving healthcare services in CKD management.