This theoretical paper aims to unravel the psychological mechanisms explaining how terrorism threat alerts can lead to public underestimation of the risk of a terrorist attack. We discuss, compare and critique a large variety of literature from psychology, public administration, and communication science. We found that throughout the world, people find threat alerts confusing. We provide an extensive psychological explanation of how confusion can lead to optimism bias and, ultimately, to underestimating terrorism threats. Literature suggests implementing three design elements to reduce the confusion that current (leveled) threat alerts create: keep it simple, call for action, and make international agreements for standardization. We discuss these practical design suggestions and give detailed directions for future research to assess the impact of terrorism threat alerts empirically.