Abstract: There is considerable evidence now that recognition of emotion from facial expression occurs far above chance, at least for primary emotions. On the other hand, not much research is available studying the process of emotion recognition. An early theory was proposed by Lipps (1907), postulating that an ‘imitation drive’ accounts for this process. According to this theory, we tend to imitate a facial expression to which we are exposed, via feedback mechanisms we realize that our own imitated facial expression is associated with an emotion, and then we attribute this emotion to the person confronting us. Using Ekman & Friesen's (1976) Pictures of Facial Affect, a study employing 20 subjects was conducted. During the first part subjects had to judge the emotions expressed in the pictures of facial affect. During this task the subjects were videotaped without their knowledge. About two weeks later the same subjects watched the video-recordings of their own expressions during the judgement task and had to judge which emotions they had decoded for the respective slides two weeks previously. Results indicate that decoding of emotions from own facial expression and decoding of the respective emotions from pictures of facial affect correspond to a degree above chance. The results are discussed with respect to the possible impact of imitation on the process of emotion recognition.