In his concise argument, ‘A relational approach to saviour siblings?’, Selgelid reiterates some of the arguments raised in the author meets critics discussion of my book, Saviour Siblings. In this response, I highlight an important misunderstanding in one of the arguments put forward by Selgelid, which forms the basis of a large portion of his analysis. Contrary to what Selgelid contends, I do not use the deafness case in my discussion of the non-identity problem to contend that the case of selecting for deafness is ethically different from the case of saviour siblings. As I state in my reply, I use the case of deafness not as a comparator for saviour siblings but rather to illustrate the different categories of risk that apply in selection cases. Given this confusion, I restate my objection to relying on the non-identity problem in evaluating risk of harm associated with the embryo biopsy process for preimplantation genetic diagnosis. Finally, I reiterate that the individual interests of saviour siblings remain important in the decision-making matrix and emphasise that Saviour Siblings offers a more contextualised approach to the welfare of the child in selective reproduction, which includes both individual and collective interests.