Diagnosis of a mental disorder is widely recognised as an important step towards effective treatment. This procedure is currently usually based on the criteria defined in the DSM-IV. Derived is a number of rating scales, used as screening tools for several disorders, among which, ADHD. These rating scales (e.g. ADHD rating scale; DuPaul, 1991; ADHD-SC4; Gadow & Sprafkin, 1997) are usually completed by parents who assess subjectively the children?s behaviour. According to our clinical experience and a plethora of research data as well (Achenbach et al., 1987; Kumpulainen et al., 1999), informants often provide different ratings of a child?s behaviour. It may happen that a child does meet or not the criteria for the diagnosis of ADHD dependent upon whether the rating scale is completed by the mother or the father. These phenomena may result from the lack of objective measures for the diagnosis of ADHD and from the overvaluation of rating scales. In our perspective, the diagnosis of ADHD should be based not only on the existence of primary symptoms, as they are described in the DSM-IV, but mainly on the evaluation of secondary difficulties that may impede the child?s functionality. Within this framework, qualitative and not quantitative assessment appears more salient. In this paper, we discuss these and other thoughts about the assessment of ADHD and we present the procedure followed in the Psychological Center of Developmental and Learning Disabilities ?ARSI?. Clinical examples are also given to illustrate the theoretical underpinnings of the procedure presented.