Abstract
This article describes the process of implementing the HCR-20 into clinical practice at two Swedish forensic psychiatric hospitals. Since the instrument is fairly new, the interrater reliability was checked in a very comprehensive way. Seven clinicians undertook the task of assessing 43 violent offenders admitted to a forensic psychiatric hospital. These assessments were done independently of each other; no discussions of the cases were allowed between the assessors. The outcome is very encouraging, with a high internal consistency and high interrater reliability overall. The whole process of implementing a research instrument into clinical practice is described, how it was done, and what experiences were found.