Background: One of the most extensively tested risk assessment instruments in offenders with an intellectual disability (OIDs) is the Violence Risk Appraisal Guide (VRAG). The purpose of this prospective study was to test the ability of this instrument to predict institutional aggression in OIDs. Method: VRAG scores were collected for 52 OIDs, and staff registered aggressive incidents during a period of 6 months. Predictive accuracy was analysed using several performance indicators. Results: The VRAG was not able to significantly predict aggressive behaviour, although a trend towards significance was found for interpersonal physical aggression. Furthermore, the VRAG was only accurate in prospectively identifying low-risk individuals. Conclusion: Compared to other studies, demonstrating good predictive validity for institutional aggression in OIDs, the VRAG performed poorly in this study. The usefulness of this instrument in identifying high- risk offenders is questioned. Until further evidence is available from studies using multiple performance indicators, clinicians have to be very careful in relying on the results of a risk assessment tool.