The assessment of children in clinical practice frequently relies on the use of assessment methods that are generally easy to put into practice. The clinician uses the information collected using testing instruments to extract specific characteristics observed in the child thus evaluated. Some instruments permit a comparison of the child's performance, in one or several areas, with that of a reference age group. Such comparisons involve the assumption that these areas operate in an isolated fashion and that any interactions with other fields may be neglected. Other instruments focus more on the identification of specific atypical signs that may be observed in a given pathology. This process aims at isolating areas of abnormal functioning and tends to neglect other dimensions that may be important to determine the global behaviour. Such approaches to evaluation present obvious drawbacks when dealing with impaired children. Indeed, these children's organisms, as well as the environment in which they have developed, possess specific characteristics that render any comparison with a normal population less relevant. One should consider the possibility that certain tasks call upon resources emanating from other areas of behaviour. From this point of view, the responses observed, typical or not, can inform us about the original means that these children put into motion in order to answer the constraints of the situation. As a consequence it becomes necessary to complete the evaluation with an approach that takes into account the reciprocal influence of various fields and that poses the question of the role of the behaviour observed for the subject's activity.