Chronic kidney disease (CKD) is a persistent medical illness which may arise due to kidney cancer or impaired renal function. Despite the fact that it is possible to halt or slow the progression of this chronic illness entirely, it is also possible to stop or decrease the course of this condition to an end stage when hemodialysis or surgical therapy is the only way to sustain the life of a patient. Timely identification and suitable treatment might enhance the probability of this occurrence. This study has explored the possibilities of several machine learning (ML) algorithms to diagnose CKD at an early stage. Initially, the research first considers 25 attributes, excluding the class attribute. However, it ultimately selects just 11 of these features as the subset for identifying CKD. Seven different ML classifiers were tested in a supervised learning environment. The experimental results revealed that the XGBoost classifier performed very well, with precision, accuracy, recall, and F1-score equivalent to 0.98. The study’s methodology suggests that current advances in machine learning provide an intriguing opportunity to discover novel ways for evaluating prediction accuracy in the context of renal disease and other similar fields.