The question I open this paper with is a simple one, how do we begin a child analysis? The answer may seem to be similarly plain, apart from the stumbling block that communication may be, we begin the same as we would with an adult analysis. But beyond communication child analysis is very particular in certain aspects and I wish to address two of these particularities in this paper, and thereby consider, what we risk when we begin a child analysis and fail to take into account what is both without and within the child. When a child begins therapy they come for a complicated number of reasons, none of which may express anything about their own desire. In addition, their continuing to attend is a decision that often lies elsewhere, with the referrer or with the parent. So we are speaking here of a demand that may operate on two or more levels but we need to add a third and that is the demand of the analytic technique whose quest is the truth, a truth which may have little to do with the perception of the 'sick' or 'problem' child. These demands can essentially be seen as lying outside of the child and yet they are nevertheless with him for every step of the treatment, they are at once 'with' him and 'out'side of him. The second aspect is that the child cannot be heard in isolation, he or she is part of a family and it is important to place the child's words in the context of his or her parent's desire. This may invoke anxiety not only on behalf of the parents but also the analyst as there may be an implicit assumption that the focus is to be the child. If we concentrate on what is 'in' the child, we risk losing the essential place of their parents words or absence of them in the child's speech, the collective nature of the 'with' that can be heard in the progress of the treatment. Through looking at the work of Mannoni and Lacan, in the first section oft his paper, I will then reconsider Freud's cases of Dora and little Hans, in the third and fourth sections, with the questions raised in mind. I will finally consider clinical material of a five month analysis with a five year old girl to explore the possible implications of these dynamics for the direction of the treatment, in the fifth and sixth sections before finally summing up the implications for a child analysis in the conclusion.