We examined the possibility of an in-group advantage in detecting intergroup anxiety. Specifically, we videotaped White and Black participants while they engaged in same-race or interrace interactions. Then we asked White and Black observers to view these videotapes (unaware of the racial context) and provide their impressions of participants' anxiety. Two results pointed to an in-group advantage in detecting intergroup anxiety. First, only same-race observers perceived a modulation of participants' anxious behavior as a function of racial context. This held true not only for relatively subjective perceptions of global anxiety, but also for perceptions of single, discrete behaviors tied to anxiety. Second, we found that only same-race observers provided descriptions of anxiety that tracked reliably with participants' cortisol changes during the task. These results suggest that White and Black Americans may have difficulty developing a sense of shared emotional experience.