It could be expected that in initial clinical interviews prompted by learning difficulties, therapists would avoid attributing the cause of the child's difficulties to a "lack of intelligence", and that the parents would avoid giving information that the therapists might interpret as an indication of a "lack of intelligence". Starting from this main assumption, we shall present the results of an analysis conducted on a series of initial clinical interviews prompted by learning difficulties. Our aim is firstly to explore whether the topic of "intelligence" is discussed in all or just some of the interviews and then to address the following main question: How is "intelligence" brought into language in this type of interview?