In recent years academic debate has (re)focused on the extent of the co-occurrence of intellectual disability (ID) and criminality, although findings from prevalence studies examining this link have been inconsistent. In April 2004 a process for transferring responsibility for commissioning health care services in UK prisons to Primary Care Trusts (PCTs) commenced. This development meant that it was important for PCTs to ascertain the need for specialist ID services in prisons in their areas. Because there were no reliable prevalence data, this research was commissioned by one such PCT.Method Using a stratified random sampling frame, data were collected from 185 young adult male prisoners aged 18 and 21 years old. Participants completed a semi structured research questionnaire, the Kaufman Brief Intelligence Test (Second Edition) (KBIT2) and the Vineland Adaptive Behavior Scales (Second Edition) (VABS2).ResultsTen per cent had an IQ composite of 69 or below, indicating a significant impairment in cognitive functioning. A further 10 per cent had IQ composite scores between 70 and 74, and 14 percent between 75 and 79. None of the sample had an Adaptive Behavior Composite (ABC) score of 69 or less, although 15 per cent scored between 70 and 79 on this measure. Whilst none of the sample could be classified as having an ID in strict diagnostic terms, four per cent scored 69 or below on the KBIT2 and in the borderline (71-79) range on the VABS2, four percent had borderline ABCs and IQs between 70 and 74; and three per cent had borderline ABCs and IQs less than 79. This equates to a point prevalence of borderline ID of 11 per cent.Conclusions: Eleven per cent of the sample had borderline ID, supporting suggestions that this group are prevalent - and easily hidden - in mainstream criminal justice settings. Differences between borderline ID inmates and non-ID counterparts are discussed, and implications for service delivery are considered.