The extent to which stigmatized interaction partners engender perceivers' threat reactions (i.e., stigma-threat hypothesis) was examined. Experiments 1 and 2 included the manipulation of stigma using facial birthmarks. Experiment 3 included manipulations of race and socioeconomic status. Threat responses were measured physiologically, behaviorally, and subjectively. Perceivers interacting with stigmatized partners exhibited cardiovascular reactivity consistent with threat and poorer performance compared with participants interacting with nonstigmatized partners, who exhibited challenge reactivity. In Experiment 3, intergroup contact moderated physiological reactivity such that participants who reported more contact with Black persons exhibited less physiological threat when interacting with them. These results support the stigma-threat hypothesis and suggest the utility of a biopsychosocial approach to the study of stigma and related constructs.