Law and Human Behavior, Vol. 15, No. 6, 1991 Psychopathy and Violent Recidivism Grant T. Harris, Marnie E. Rice, and Catherine A. Cormier* The violent recidivism rates of 169 adult male mentally disordered offenders released from a maximum security psychiatric hospital were compared over an average 10-year follow-up period. Forty percent of the total and 77% of the psychopaths (as defined by the Psychopathy Checklist) committed a violent offense. It was possible to predict outcome with considerable accuracy using combinations of child- hood history, adult history, index offense, and institutional or program variables. However, the Psychopathy Checklist alone performed at least as well as any combination of variables and also improved upon the prediction based on criminal history variables. Psychopaths continued to recidivate at a higher rate than nonpsychopaths even beyond age 40. Persons diagnosed as psychopaths occupy large numbers of beds in correctional and mental health facilities. Wong (1984) found that 15%--30% of a sample of federal inmates met the commonly used research criterion for the diagnosis of psychopathy, depending upon the security level of the correctional institution. There is evidence that psychopaths commit disproportionately high numbers of criminal and violent offenses (Hare & Jutai, 1983; Hare & McPherson, 1984; McCord, 1982). Although there is general agreement about the salient characteristics of the psychopath (Buss, 1966; Cleckley, 1976; Hare, 1970), a key problem in the eval- uation of the literature on psychopathy relates to its definition. It is difficult to evaluate the evidence linking psychopathy and violent criminality because crim- inal behavior is almost invariably a defining property of the term psychopathy and the closely related term, antisocial personality disorder (American Psychiatric Association, 1980; Hare, 1970; Hare & McPherson, 1984; Spitzer, Endicott, & * This research was supported by the Ontario Mental Health Foundation under grant No. 952-86-88. Thanks are due to R. Hare and V. Quinsey for comments on earlier versions of this paper and to V. Bell, C. Lang, L. Koshkerian, S. Robinson, and C. MacKnight for collecting the data. Corre- spondence should be sent to the authors at the Mental Health Centre, Penetanguishene, Ontario, L0K 1P0. 625 0147-7307/91/1200-0625506.50/0 9 1991 Plenum Publishing Corporation