Background Among mainstream offenders, the severe personality disorder of psychopathy has considerable importance as a construct. The disorder has long been associated with failure to make treatment progress. Previous work has identified that psychopathy as a disorder occurs in samples of offenders with intellectual disability (ID), and suggests that the Psychopathy Checklist – Revised (PCL‐R: Hare, 1991, 2003) as a measure of the disorder has adequate reliability and validity (Morrissey et al., 2005). The present study aimed to compare the predictive power of the PCL‐R in relation to treatment progress with a more general assessment of violence risk, the HCR‐20 (Webster, Douglas, Eaves, & Hart, 1997).Method A sample of 73 residents in a high security intellectual disability service, who had previously been assessed using the PCL‐R and the HCR‐20, were followed up at 2 years post‐assessment, and their outcome determined in terms of two distinct dichotomous variables reflecting definite positive treatment progress and definite negative treatment progress respectively.Results In line with predictions, the PCL‐R Total score and Factor 1 score (Interpersonal and Affective aspects of psychopathy) and the HCR‐20 Total score were significantly inversely associated with a positive move from high to medium security hospital conditions within 2 years of assessment. However against prediction, the PCL‐R Total score had incremental validity over the HCR‐20. The PCL‐R Total and Factor 1, but not the HCR‐20 Total score, were also significantly associated with negative treatment progress in terms of a move to more restricted treatment conditions.Conclusion Psychopathy, and in particular its interpersonal and affective manifestations, is a construct which appears to be associated with indirect measures of treatment progress in this high security ID group. However, caution should be applied in the use of a construct with potentially negative connotations in an already devalued population.