This experiment examined how contextual cues can induce stereotype threat and reduce one's sense of belonging. Male and female undergraduates majoring in math, science, and engineering viewed a videotape of a discussion that supposedly took place at a conference on leadership in the sciences, assessing whether the conference should be hosted by the local university in the future. The gender composition of the discussants was varied so that either 75% of the discussants were males (reflecting approximate sex differences in obtaining degrees in these fields in the U.S.), or 50% males and 50% females. Physiological measures of arousal were collected while the students watched the videotape, and later, memory of details in the video and the experimental setting were assessed. In terms of memory performance, women who viewed the video comprised mostly of men recalled more details about the video and showed greater recall of science and math items that had been present in the experimental setting. Woman in this condition also showed greater skin conductance, decreases in heartbeat intervals, and greater sympathetic activation of the cardiovascular system than did women who watched a gender-balanced video. These women also reported that they would feel less comfortable attending the conference and would be less likely to do so compared with women who watched the gender-balanced discussion. Men did not differ on any of these measures as a function of the video they watched, although they did report greater interest in attending the conference if it was represented as gender-balanced. These results suggest that expected gender minority status is sufficient to produce stereotype threat as reflected in attentive vigilance, physiological arousal, and a reduced sense of belonging.