Understanding how self-representation is built, maintained and updated across the lifespan is a fundamental challenge for cognitive psychology and neuroscience. Studies demonstrate that the detection of body-related multisensory congruency builds bodily and facial self-representations that are crucial to developing self-recognition. Studies showing that the bodily self is more malleable than previously believed were mainly concerned with full-bodies and non-facial body parts. Crucially, however, intriguing recent evidence indicates that simple experimental manipulations could even affect self-face representation that has long been considered a stable construct impervious to change. In this review, we discuss how Interpersonal Multisensory Stimulation (IMS) paradigms can be used to temporarily induce Enfacement, i.e., the subjective illusion of looking at oneself in the mirror when in fact looking at another person's face. We show that Enfacement is a subtle but robust phenomenon occurring in a variety of experimental conditions and assessed by multiple explicit and implicit measures. We critically discuss recent findings on i) the role of sensory extero/proprio-ceptive (visual, tactile, and motor) and interoceptive (cardiac) signals in self-face plasticity, ii) the importance of multisensory integration mechanisms for the bodily self, and iii) the neural network related to IMS-driven changes in self-other face processing, within the predictive coding theoretical framework.