Research on deception detection has established that, on average, accuracy of deception judgments is slightly above chance level. In addition, judges generally exhibit a truth bias (i.e., an overestimation of truthful messages). In the present research, we will investigate whether these findings are moderated by judges' and suspects' racial group. In three experiments, participants will be asked to judge the veracity of Black and White targets who tell the truth or lie. We hypothesize that accuracy will be lower in cross-race judgments compared to own-race judgments.