Current discussions on the concept of agency in the inter- and multidisciplinary field of Childhood Studies are a reaction to the deficits revealed in the way the New Sociology of Childhood has viewed agency: as an attempt to gain a more nuanced, differentiated understanding of agency on both an empirical and a theoretical level. My intention here is to contribute to the current reconceptualisation of agency, but not in the sense of providing a better or more nuanced understanding of individual or human agency. Quite the contrary, I argue for a different understanding of agency as social and collective, which allows for different sensitivities and methodologies in research on childhood. I will bring into play selected contributions on agency in the social theory and social anthropology of the last two decades that share a relational/ relativistic approach towards the social. The core argument that I want to push forward is that a relational conception of agency can be one productive reaction towards the claim by Prout and others that we need to reflect on existing understandings of agency in the field, all the time striving for a qualitatively different approach toward agency. This perspective turns away from predominantly intentional and cognitive understandings of agency. Hence, agents are not substantialised agents, but often consist of overlapping entities or fabrics, which are complex and in motion. Consequently, agency can be seen as a realised, situated and permuted capacity, which can be accomplished through the combination of various interconnected “persons” and “things”.