Internationally there is growing enthusiasm and expectations for researchers to make their research data openly available for use by other scholars and interested parties, including the wider public. Funding organisations are increasingly introducing requirements for the publication of datasets and actively encouraging data sharing; concurrently, institutions and researchers are looking at new ways of storing, managing and disseminating their research data. There is a corresponding interest in experimenting methodologically with data sharing and addressing its ethical, methodological and practical challenges. Secondary analysis of social science data has been more typically associated with re-analysis of quantitative datasets, either using a new technique of analysis or asking questions on a topic that was not part of the original study’s main focus. Some protocols for qualitative reanalysis can be learnt from the experience of quantitative secondary analysis, but there are obvious and well-documented differences in the type of data and process for eliciting or collecting data, in negotiating the context of the original study and in how data are recontextualised in any subsequent interpretation. This paper explores directions and dilemmas in the archiving and sharing of qualitative research, taking a specific focus on studies of childhood, education and youth, predominantly from across the social sciences. It reports on a program of work conducted in partnership with the Australian Data Archive to develop a new website and community of practice, Studies of Childhood, Education & Youth (SOCEY) (see https://socey.hasscloud.net.au). This website provides a portal to a new pilot archival repository of qualitative research project data. Six projects have been archived to date and the paper reports on the practical, ethical and methodological dilemmas we encountered in developing the new repository and in archiving the projects housed within it. We argue for more critical attention to the double-edged affordances and ambivalent effects of data sharing and openness and to how data archives are imagined, constructed and curated.