This article examines the processes that contribute to the stigmatization of a group of people typically identified as “children in care” or “looked after children.” In particular, we will look at the ways that we (adults, professionals, and carers) interact with these children, based on their status as both children and members of a socially marginalized and disadvantaged group, and how these modes of interaction can inhibit dialogue—a dialogue that is needed if we are to base our conceptions regarding the needs of these children on a more accurate understanding of their experiences and perspective. The problem is particularly challenging because the very terminology we use in the care community to identify this group is a product of the damaging preconceptions that have affected our interactions with its members and, we argue, it serves to reinforce those preconceptions. Using Fricker's work on epistemic injustice, in conjunction with evidence regarding how accusations of abuse and neglect of these children have been addressed in numerous cases, we illustrate the problems we have in hearing the voices of members of this group and the harmful effects this has on their own ability to understand and articulate their experiences. These problems represent “barriers to disclosure” that need to be surmounted if we are to establish a more inclusive dialogue. Currently, dialogue between these children and those of us charged to “look after” them is too often characterized by a lack of trust: not only in terms of the children feeling that their word is not taken seriously, that their claims are not likely to be believed, but also in their feeling that they cannot trust those to whom they might disclose abuse or neglect. The goals of the paper are modest in that we aim simply to open up the debate on how to meet this epistemic challenge, noting that there are specific problems that extend beyond those already identified for hearing the voices of other victims of epistemic injustice. Explicitly recognizing the nature and extent of the problem still leaves us a long way from its solution, but it is a crucial start.