This paper presents a study on detecting phishing websites using machine learning techniques. The research aims to identify the best classifier and feature selection method for phishing detection. The study evaluates 24 classifiers from six learning strategies and four feature selection methods using two phishing datasets. The results show that the RandomForest, FilteredClassifier, and J-48 classifiers perform best in detecting phishing websites. Additionally, the InfoGainAttributeEval method is found to be the best feature selection method. The study also highlights that the trees learning strategy performs best in terms of accuracy, true positive rate, and F1-score, while the rules learning strategy performs the worst. The findings suggest that the RandomForest classifier is more suitable for datasets with a higher number of class labels and lower number of instances and features. The InfoGainAttributeEval method is recommended for feature selection in phishing detection. The study concludes that machine learning techniques are effective in detecting phishing websites and suggests future research directions, including the use of ensemble models and metaheuristic algorithms for feature selection.This paper presents a study on detecting phishing websites using machine learning techniques. The research aims to identify the best classifier and feature selection method for phishing detection. The study evaluates 24 classifiers from six learning strategies and four feature selection methods using two phishing datasets. The results show that the RandomForest, FilteredClassifier, and J-48 classifiers perform best in detecting phishing websites. Additionally, the InfoGainAttributeEval method is found to be the best feature selection method. The study also highlights that the trees learning strategy performs best in terms of accuracy, true positive rate, and F1-score, while the rules learning strategy performs the worst. The findings suggest that the RandomForest classifier is more suitable for datasets with a higher number of class labels and lower number of instances and features. The InfoGainAttributeEval method is recommended for feature selection in phishing detection. The study concludes that machine learning techniques are effective in detecting phishing websites and suggests future research directions, including the use of ensemble models and metaheuristic algorithms for feature selection.