Although intergroup friendships have been shown to reduce prejudice, little research has considered whether interventions fostering intergroup friendship would be effective in highly prejudicial contexts. We conducted a quasi‐experiment (N = 61) to test whether a contact‐based intervention based on intergroup friendship could reduce bias against Roma people among non‐Roma Hungarians. Participants in the contact condition engaged in a face‐to‐face interaction with a Roma person, and responded to questions involving mutual self‐disclosure. Through pre‐ and post‐test questionnaires, we observed significant positive change in attitudes and contact intentions among participants in the contact condition, while these effects were not observed among participants in the control condition. Positive change was moderated by perceived institutional norms, which corroborates the potential of contact‐based interventions.