One of the most critical challenges facing banking institutions is customer churn, as it dramatically affects a bank’s profits and reputation. Therefore, banks use customer churn forecasting methods when selecting the necessary measures to reduce the impact of this problem. This study applied data mining techniques to predict customer churn in the banking sector using three different classification algorithms, namely: decision tree (J48), random forest (RF), and neural network (MLP) using WEKA. The Results showed that J48 had an overall superior performance over the five performance measures, compared to other algorithms using the 10-fold cross-validation. Additionally, the InfoGain and correlation features selection method was used to identify significant features to predict customer churn. The experiment revealed that both algorithms work better when all features are utilized. In short, the results obtained help in predicting which customer is likely to leave the bank. Furthermore, identifying these essential features will help banks keep customers from churning out and compete with rival banks.