Emotions are essential constituents of the meaning and quality of human existence. Even though a generally accepted definition of emotions does not exist, emotions are widely held nowadays to be made up of three distinct components: the subjective, the expressive and the neural component. Surprisingly, there is a substantial lack of studies that have simultaneously examined all three components of emotion processing.In this dissertation project, all three components were simultaneously measured while emotions were elicited. The central question was whether peripheral feedback from facial muscles modulates subjective experience and neural activity within central emotional brain regions. In the first study, two methods, facial electromyography (EMG) and the Facial Action Coding System (FACS) for measuring the expressive component were tested. In the second study, the subjective component (feeling ratings), the expressive component (EMG), and the neural component, functional magnetic resonance imaging (fMRI) were simultaneously measured while participants perceived emotional scenes and facial expressions and imitated facial expressions of different basic emotions. For a direct test whether peripheral feedback from emotion-specific facial muscles modulates neural activity, the EMG data was modelled parametrically.Results of the first study indicate that EMG is an adequate method for measuring the expressive component since it dissociates basic emotions and additionally detects non-visible facial movements. Results of the second study indicate that expressive behavior and subjective experience are positively correlated and that imitation of emotional facial expressions increases neural activity in emotion-relevant brain regions. Specifically, compared with the mere observation of emotional facial expressions, the imitation of happy, fearful, sad, disgusted and angry facial expressions significantly increased neural activity in the amygdala. These findings suggest that facial feedback modulates neural activity in central limbic areas during the intentional imitation of facial expressions. Furthermore, emotion-specific EMG muscles correlated with amygdala activity while observing positive emotional scenes and while observing and imitating happy faces, but not while observing or imitating any negative stimuli.Results are discussed within the framework of the facial feedback hypothesis which assumes a functional link between expression and subjective experience and the notion for a facial feedback mechanism on the neural level is supported, in particular for positive emotions.