Two experiments explored the role of mimicry and self-perception processes in emotional contagion. In Study 1, 46 subjects watched two brief film clips depicting an episode of startled fear. In a separate procedure, subjects adopted facial expressions of emotion, and reported whether the expressions had caused them to feel corresponding emotions. Those who reported feeling the emotions were identified as more responsive to self-produced cues for feeling. Subjects who visibly moved to mimic the behavior of the actor were significantly more likely to be those who were more responsive to self-produced cues. In Study 2, 57 subjects watched three film clips depicting happy people. During clips when they inhibited the movements of their faces, subjects reported less happiness than during clips when they moved naturally and were able to mimic, or when they exaggerated their movements. This effect occurred only among subjects who, in a separate procedure, had been identified as more responsive to self-produced cues.