A set of studies was designed to assess gender and racial emotion stereotypes and to investigate the effect of stereotyped holdout jurors expressing emotion on their ability to exert minority influence during deliberation. In a deception paradigm, participants were told that they were engaged in a computer-mediated deliberation with five other mock jurors about a murder case, when in reality they were reading a pre-written fictional deliberation script. After mock jurors reported their initial verdict preference, all participants then saw the same false pre-determined feedback in which one holdout juror always argued for the verdict opposite of the participant’s original verdict. The holdout expressed either no emotion, anger, or fear; and was either a man or a woman (Study 2), or a White or Black man (Study 3). When holdouts expressed no emotion or fear the participants exhibited no attitude change (Studies 2 and 3). After a man holdout expressed anger, however, participants began to doubt their original opinion more (i.e., the holdout exerted minority influence); whereas after a woman expressed anger, participants became more confident in their original opinion (Study 2). Anger expression did not affect participants’ confidence in Study 3—regardless of holdout race. These studies increase understanding of how historically underrepresented jurors might decrease their credibility and potential to exert minority influence through expressing emotion on juries. These studies also have important implications for diversity in persuasion and group decision making, particularly in the jury context.