We thank Patricia Hogan for her comments challenging current definitions of intellectual disability and highlighting the difficulty of accurate assessment of psychotic illness in individuals with intellectual disability. With respect to the former, we note the importance of applying standard definitions and nomenclature in the study of the epidemiology of dual diagnosis. The criteria used to define intellectual disability affect prevalence rates and the use of IQ criteria alone rather than the dual criteria of IQ and adaptive behaviours will have a marked impact on rates. ^{Reference Whitaker1} We employed the American Association on Mental Retardation dual criteria in our study. The use of dual criteria is the most common approach across services and in research, and is consistent with DSM–IV and ICD–10 definitions. As the American Association on Mental Retardation criteria are the basis of service eligibility in Western Australia, their use ensures a thorough assessment of individuals on the intellectual disability register and greater confidence that cases have been correctly classified in this study. Although the difficulty of diagnosing psychosis accurately in individuals with intellectual disability is well documented, ^{Reference Deb and Weston2,Reference Friedlander and Donnelly3} our paper highlights another pressing issue. The poor recognition of dual diagnosis in affected individuals as a result of the administrative separation between intellectual disability and mental health services has led to a serious underestimate of the prevalence of dual diagnosis and has created structural impediments to inter-agency approaches to integrated, person-oriented clinical practice. Critical improvements are needed both in the structure of service provision and in clinical education programmes to ensure dual diagnosis is correctly identified and appropriately treated. ^{Reference Bouras and Holt4,Reference Catinari, Vass, Ermilov and Heresco-Levy5} Otherwise dual diagnosis will continue to be recognised and treated ineffectively or, at worst, missed altogether, with important implications for best practice.