["Legal and Criminological Psychology, EarlyView. ", "\nAbstract\n\nPurpose\nThe present study examined whether three commonly used assessment instruments can predict treatment non‐completion from a community treatment programme for individuals convicted of sexual offences.\n\n\nMethod\nAll 289 male sexual offenders in the sample completed a standard intake assessment, which included a clinical interview, completion of the URICA, the Static‐99R and the Stable‐2007. Clinicians certified in the administration of the Static‐99R and Stable‐2007 administered, scored and interpreted these instruments based on information from the clinical interview and collateral records.\n\n\nResults\nThe Stable‐2007 consistently predicted treatment attrition, whereas the Static‐99R and URICA did not add significant predictive power once dynamic risk was accounted for. Item‐level analysis of the Stable‐2007 revealed that Poor Problem Solving was the strongest individual risk factor associated with treatment non‐completion, though other subscales of the Stable‐2007 also predicted non‐completion.\n\n\nConclusions\nThese findings suggest that dynamic criminogenic needs may play a central role in identifying which sexual offenders are at risk of not completing treatment. The implications of this exploratory study are discussed, including how the Stable‐2007 can help identify criminogenic needs for intervention to improve treatment completion rates.\n\n"]