People in social media post massive amounts of different types of data including text messages, photos, and links. They share their personal opinions, feelings, and even their health status. The high volume of health-related tweets can be used as a tool to track the activities of different infectious diseases. In this paper, we describe our work to process Twitter data to detect self-reported cases of the flu using supervised machine learning methods. The results obtained on a large set of tweets posted in English during the winter season prove that machine learning classifiers are effective in detecting possible self-reported flu cases.