The present study provides a comparison of face-to-face and mediated hurtful communication in close relationships. Drawing on previous studies on hurtful communication and computer-mediated communication (CMC), an escalating hypothesis was posited that mediated hurtful messages would be perceived as more controllable, intentional, and hurtful than face-to-face (FtF) hurtful messages. Study 1 tested these predictions. Survey responses from college students who were randomly assigned to report either mediated or face-to-face hurtful interaction with a friend or romantic partner confirmed higher perceived controllability (i.e., being more deliberate on crafting hurtful messages) in the CMC than the FtF condition. Although intent and hurt were not different between the two contexts in the full sample, higher intent was found in CMC than FtF in romantic relationships. Thus, Study 2 was conducted, focusing on a comparison of FtF and text messaging in romantic relationships. Also, perceived face threat and relationship aspects—distancing effect of hurtful interactions and the effect of relationship satisfaction—were assessed. Face threat was posited to be lower in CMC than FtF context since Study 1 suggested that self-focused appraisals (e.g., humiliation) were lower for mediated hurtful messages. This difference in face threat was considered to result in similar levels of intent and hurt between the two contexts, although perceived controllability is higher in CMC. Results of Study 2 confirmed higher deliberation in CMC but failed to confirm higher intent or hurt in CMC. Further, face threat was not different between the two contexts, and controlling for face threat did not reveal the escalating effect (i.e., higher intent or hurt in CMC). Regarding relationship aspects, higher satisfaction and lower intent were associated with less distancing effect, and relationship satisfaction was negatively related to deliberation, intent, hurt, and distancing. These tendencies were not different between CMC and FtF contexts. These results suggest that mediated hurtful communication is a complex phenomenon in which various factors should be considered. The implications of these results were discussed, and suggestions for future studies were also offered. / text