With mounting evidence of the harmful societal consequences of affective polarization, it is crucial to find ways of addressing it. Employing a randomized controlled trial, this study tested the effectiveness of an intervention based on theories of intergroup contact and interpersonal communication in reducing affective polarization in the context of Brexit. Participants were 120 UK self-identified Leavers and Remainers. Sixty Leaver-Remainer dyads were randomized to engage in either a facilitated intergroup interaction or a control interaction, which was equivalent in structure and tone but was unrelated to Brexit identities. Different aspects of affective polarization were assessed one month prior, immediately after, and one month after the intervention. Results indicate that the intervention increased warmth toward the outgroup, reduced unfavourable attributions of the sources of outgroup positions, and increased willingness to compromise, but only short-term. There were no statistically significant longer-term effects of the intervention. Evidence of selective attrition further suggests that those with more extreme baseline opinions were more likely to drop out. Our findings highlight the challenges of designing effective interventions that engender enduring attitude change in polarized contexts and of engaging those with extreme political views. This study can provide a useful framework for future research.