The authors observe that the risk assessment of offenders occurs within a legal context where issues such as societal expectations, personal responsibility, and judicial decisions intersect. Resolving the ethical issues pertaining to offender assessment is dependent upon acknowledging this legal context. In making this argument, the authors review current violence and sexual risk assessment practices in correctional settings and propose four central research questions for risk assessment: (1) what are the best criteria for choosing a risk-assessment instrument; (2) can the accuracy of actuarial instruments by improved; (3) what is the best way to combine risk factors in a global assessment; and (4) how can we tell if treatment has reduced recidivism risk for an individual offender? Next, risk assessment based on static and dynamic variables are reviewed along with multi-wave, prospective studies of dynamic risk factors. Future directions for research on dynamic risk are proposed and include the questions of whether dynamic variables add to static measures and questions regarding the nature of dynamic variables. Postrelease risk management strategies are reviewed, which vary considerably from risk-assessment-based strategies to political legislation. Researchers are challenged to affect public policy concerning risk management, particularly concerning the risk management of sexual offenders. The most pressing research questions in this regard are identified and include questions about the treatment components that lead to a reduction in sex offender recidivism, questions about which dynamic risk factors are most meaningful for public policy, and questions about the standards for enacting a risk-management model. Finally, methodological issues concerning violent and sexual risk assessment are discussed and critiqued for their general lack of empirical soundness. References