Phishing attack, one of the commonly encountered cybersecurity threats, employs deceiving tactics to steal sensitive information of an individual like his bank account passwords, email passwords, or any other detail required for performing financial transactions. They lure victims into their trap by generating fraudulent emails or legitimate websites and convince them to share their personal details. With advancement of technologies and devices, their attack becomes more prominent and realistic. To deal with such powerful adversary, we must be able to identify and categorize them based on their severity so as to flag those websites. So, in our work, we have employed some machine learning algorithms to successfully those websites affected by phishing attack. The algorithms under consideration are logistic regression, gradient boosting, Naïve Bayes, and decision tree. Our experimental analysis clearly indicates that random forest provides as with best evaluation performance metrics parameters over rest of the machine learning algorithms in terms of accuracy, recall, precision, and F1 score.