Research on offenders with intellectual disabilities (IDs) in the criminal justice arena is on the rise, reflected by a growing number of relevant publications each year. However, there is a long recognized methodological problem that hampers the comparability of empirical studies and that raises doubts about the accuracy of prevalence rates, comorbidities, and various correlates and characteristics. In this paper we will argue that the crux of the problem can, on the one hand, be found in the plurality of assessment methods for intelligence and adaptive functioning, which are not all sufficiently reliable and valid. On the other hand, assessment of IQ in criminal justice and mental health-related areas appears to be informed more by practical aspects and needs rather than grounded in a solid theoretical model. Hence, we suggest that the Cattell–Horn–Carroll (CHC) model of intelligence has potential value in this regard, and deserves a closer look. Finally, we will discuss its incorporation into, and possible implications for, criminal justice practice and future study designs. Copyright © 2012 John Wiley & Sons, Ltd.