Two experiments (N= 443) were conducted to investigate the effects of a defendant's emotion level during testimony on mock jurors' decisions. In Experiment 1, the defendant's level of emotion (low, moderate, high) and mode of presentation (audio, video) were varied. The defendant displaying a low level, as opposed to a higher level of emotion was perceived as more guilty and less credible. In Experiment 2, using only the video mode, emotion level and evidence strength (strong, weak) were varied. Defendant emotion level tended to affect jurors' decisions only when the evidence against the defendant was weak (i.e., a stronger display of emotion was associated with a lower proportion of guilty verdicts, shorter sentence assignments, and perceptions of a more honest defendant). Path analyses for both experiments indicate that the effects of emotion on perceived guilt level are mediated by perceptions of the defendant (e.g., the defendant's level of honesty). Implications of using defendant emotion level for determining guilt are discussed.