Abstract As a result of recent legislative changes in both the United States and Canada regarding sexually violent predators (Glancy, Regehr, 2001; Zonona, 1999), mental health practitioners are increasingly being called upon to provide predictions regarding the future dangerousness of convicted sex offenders. Prediction of dangerousness based solely on clinical assessments for offenders of any kind and in particular sexual offenders, however, has proved to be remarkably inaccurate and to result in very low interrater reliability between professional assessors (Hilton, Simmons, 2001). Consequently, there has been considerable effort in the past decade to develop actuarial tools with the aim of improving predictive accuracy. Developers of the tools have reported favorable results in terms of predictive validity, but, nevertheless, considerable controversy exists about the place of actuarial testing in the assessment of sexual offenders (Zonona, 2000; Sreenivasan, Kirkish, Garrick, Wineberger, Phenixa, 2000). The original actuarial instruments focused exclusively on “static” or historical factors such as the age at first offence and the nature of violent offenses. The developers suggested that these tools for the prediction of dangerousness are accurate enough to be used in isolation and that adjunctive clinical assessments not only may fail to add to the predictive validity but in fact may be detrimental (Quinsey, Khanna,, Malcolm, 1998).