In this study the factorial validity and the reliability of DSM-IV related ADHD symptom ratings made by care professionals working in residential treatment centres were determined in a sample of 412 residential youngsters. Three concurrent models of the ADHD disorder were investigated, a one-factor model comprising all 18 symptoms, a two-factor model with the Inattention and Hyperactivity/Impulsivity symptoms, respectively, combined and a three-factor model comprising Inattention, Hyperactivity, and Impulsivity symptoms, respectively. An analysis of the covariance structure shows acceptable fits for both the two- and the three-factor models, slightly favouring the three-factor model. The internal consistencies, the test-retest reliabilities, and the inter-rater reliabilities turned out to be good to excellent for all scales based on each of the three concurrent models.