Online health forums are areas of exchange where patients, on condition of anonymity, can speak freely on their personal experiences. These resources are a gold mine for health professionals—giving them access to patient to patient, patient to health professional and even health professional to health professional exchanges. In this study, we used text mining techniques to analyse health forums in order to extract emotions (e.g., joy, anger, surprise, etc.) expressed by patients. After a study of real messages, we demonstrate the difficulty of manual annotation due to the low level of agreement between humans. We propose a method to identify the polarity of a message and extract one or several emotions. This method was validated on a substantial real dataset.