This paper addresses the issue of phishing detection using machine learning techniques. The primary objectives are to identify the most effective classifier among twenty-four different classifiers representing six learning strategies and to determine the best feature selection method for phishing datasets. Two datasets, one binary and one multiclass, were used for evaluation, and eight metrics were considered for performance assessment. The results indicate that the RandomForest, FilteredClassifier, and J-48 classifiers perform best in detecting phishing websites. Additionally, the InfoGainAttributeEval method was found to be the most effective among four feature selection methods. The study also highlights the importance of feature selection in improving classification accuracy and suggests that ensemble models incorporating the top three classifiers could be a promising approach for future research. The paper concludes by recommending the use of metaheuristic algorithms for designing more efficient feature selection methods.This paper addresses the issue of phishing detection using machine learning techniques. The primary objectives are to identify the most effective classifier among twenty-four different classifiers representing six learning strategies and to determine the best feature selection method for phishing datasets. Two datasets, one binary and one multiclass, were used for evaluation, and eight metrics were considered for performance assessment. The results indicate that the RandomForest, FilteredClassifier, and J-48 classifiers perform best in detecting phishing websites. Additionally, the InfoGainAttributeEval method was found to be the most effective among four feature selection methods. The study also highlights the importance of feature selection in improving classification accuracy and suggests that ensemble models incorporating the top three classifiers could be a promising approach for future research. The paper concludes by recommending the use of metaheuristic algorithms for designing more efficient feature selection methods.