This article is one of the first to examine the relationship between criminal activity and the influence on it over time by mental illness. The authors discuss: how often mentally ill offenders commit crimes motivated by psychiatric symptoms; legal and research definitions of direct relationships; difficulties in distinguishing between symptoms and traits; how consistent the relationship between criminal behavior and mental illness is over time-the issue of "direct crimes"; legal and research definitions of the consistency of direct relationships; and the study's implications. "[P]rograms will be most effective in reducing recidivism if they expand beyond psychiatric symptoms to address strong variable risk factors for crime like antisocial traits" (p. 439).