Prior studies have identified an “enfacement effect” in which participants incorporate another’s face into their self-face representation after observing that face touched repeatedly in synchrony with touch on their own face (Sforza et al., 2010; Tsakiris, 2008). The degree of self-face/other-face merging is positively correlated with participants’ trait-level empathy scores (Sforza et al., 2010) and affects judgments of the other’s personality (Paladino et al., 2010), suggesting that enfacement also modulates higher-order representations of “self” and “other” involved in social and emotional evaluations. To test this hypothesis, we varied not only whether visuo-tactile stimulation was synchronous or asynchronous but also whether the person being touched in the video displayed an emotional expression indicative of threat, either fear or anger. We hypothesized that participants would incorporate the faces of fearful others more than the faces of angry others after a shared visuo-tactile experience because of a potentially stronger representation of the sight of fear in somatosensory cortices compared to the sight of anger (Cardini et al., 2012). Instead, we found that the enfacement effect (i.e. greater self-face/other-face merging following synchronous compared to asynchronous visuo-tactile stimulation) was abolished if the other person displayed fear but remained if they expressed anger. This nonetheless suggests that enfacement operates on an evaluative self-representation as well as a physical one because the effect changes with the emotional content of the other’s face. Further research into the neural mechanism behind the enfacement effect is needed to determine why sight of fear diminishes it rather than enhancing it.