Chronic kidney disease (CKD) significantly burdens patients, progressively damaging their kidneys and hindering waste removal. This leads to a buildup of toxins, eventually causing kidney failure. CKD not only decreases life expectancy but also diminishes the overall well-being of those affected. CKD poses a growing health challenge due to its increasing prevalence. Catching chronic kidney disease (CKD) early is key to managing it effectively. This research explores the potential of machine learning algorithms for predicting the onset of chronic kidney disease. We employed Python libraries in Jupyter Notebook to analyze two publicly available CKD datasets. Extensive preprocessing on the first dataset included Exploratory Data Analysis (EDA), random value imputation, data cleaning, and visualization techniques. Feature encoding was applied before evaluating ten machine learning models (Gradient Boosting, Random Forest, etc.). Didn’t have to do anything on the second dataset because it was clean and preprocessed. Both datasets were assessed using training and testing accuracy, confusion matrix for prediction details, weighted precision, recall, and F1-score. Gradient Boosting, XGBoost, Random Forest, and Extra Trees achieved high accuracy (above 0.975) in both datasets. While random value imputation for the missing values caused slight variations in the first dataset, the top-performing models remained consistent. By comparing different machine learning models, this study highlights their potential for predicting chronic kidney disease and identifies the most effective ones, opening doors for further exploration.