AbstractJudgments about people’s trustworthiness are made frequently and have important real-life consequences. However, the accuracy of these judgments is debated. We therefore systematically reviewed the current evidence for accurate trustworthiness detection in the literature. The overall evidence for accuracy is rather mixed; although we find only limited evidence for accurate trustworthiness detection from neutral photographs, trustworthiness detection becomes more accurate when the rater and target interact, when the target presentation resembles face-to-face contact, and when the target presentations contain cues or signals about the target’s trustworthiness. We also find that the current literature lacks an overarching research agenda, which leads to a large heterogeneity in the extant studies’ operationalizations. We address some of these operationalizations and suggest the following guidelines for future research: Studies should engage in stronger theory building, experimentally test moderators, strengthen generalizability by recruiting large target pools, and use appropriate methods for the analysis of nonindependent data.