Risk assessment instruments for violence and reoffending are widely used throughout the world. According to researchers, there are many different reasons to use these instruments; for instance, they are thought to reduce violence, save money, and improve treatment planning. In this article, we create a taxonomy to classify these risk management outcomes into agency, professional practice, and evaluee domains. Through a review of research, we show that instruments do not always achieve their goals. First, agencies encounter problems in successfully implementing instruments. Second, a lack of follow through can occur between risk assessments and the subsequent phases of risk management, such as case planning and intervention delivery. By drawing from the field of implementation science, we create an agenda for research.