The sample consisted of all offenders with a history of violent crime incarcerated at some time during the study period (October 1999-June 2002) at the A-ward in the Swedish maximum-security prison Hall (n=47). These inmates were assessed with the HCR-20 risk assessment scheme, Version 2. Approximately one-third of the study group were administered a second HCR-20 assessment after a minimum of 3 months and a maximum of 24 months, so as to assess the possible effects of interventions. The assessments were followed by discussions with staff members that produced risk management strategies for each inmate. Through these discussions, staff members became aware of each inmate's personality characteristics, his risk factors for violence, and how best to manage the risk factors. The follow-up study showed no significant decrease in important risk factors for violence in the sample; however, the number of violent incidents decreased significantly during the study period. The findings suggest that even though risk factors for violence may persist in individual inmates, violent behavior can be reduced through strategically applied risk-management policies and mechanisms. The study also supported the theoretical assumption that changes in risk factors are more likely in some populations (e.g., general psychiatric) than in others (e.g., correctional) depending on the characteristics of the study group and the risk factors involved. This is an important consideration when conducting evaluation research with risk assessment instruments. 2 tables, 4 figures, and 26 references