Different research studies have proved that emotions meet a pivotal role in cognitive processes and in particular the studies made by Damasio who argues that human-beings without emotions could not make the simplest decision (Damasio 1994). We think that the fail of Intelligent Distance Learning Systems to achieve an efficient learning is mainly resulting from the lacking of Emotional Intelligence abilities. These systems require a capacity to manage the emotional state of the learner so as to be in the best conditions for learning. To achieve this goal, it is very important to anticipate the emotional response of the learner after the happening of an event in the learning session. In this paper, we propose a method for predicting the learners' emotional response by using an intelligent agent called ERPA (Emotional Response Predictor Agent). This agent uses a case-based reasoning, an Artificial Intelligence technique, and a Learner's Event-Appraisal Model.