How does ingroup-outgroup contact reduce social and even political polarization? Despite the large number of studies on the well-known contact hypothesis, the current literature lacks experimental field evidence or even behavioral results. In this project, I examine this connection in an online environment, a venue where polarizations are particularly extreme. In collaboration with a non-governmental organization, English-speaking participants are randomly assigned to brief Zoom conversations with a person from a socially stigmatized group. The conversation will focus on one group membership, e.g., being Jewish, but other prejudiced group memberships may be indirectly reduced (e.g., female versus male Jews). The research component aims to examine how this contact intervention reduces prejudice by examining mechanisms of group identification, empathy, and a time-lagged measure of behavior change.