In courts across the country, judges depend on mental health experts to determine whether mentally disordered people are dangerous. But experts' ability to predict violence is severely limited, and they are wrong as often as they are right. This study reviews two decades of research on mental disorder and offers new empirical and theoretical work that will pave the way for more accurate predictions of violent behavior.

"Essential for all those who are interested in the study of risk assessment of violence. It is particularly important for the researcher in this area. . . . For the clinician who must make violence assessments it is important reading as well."—Stewart Levine, Bulletin of the American Academy of Psychiatry and the Law