Advances in computational capacity, the accelerating rate at which digital research data are accumulating, and the associated opportunities for new ways of doing data-driven research have brought increased attention to issues of data access, integrity, and preservation. Funders, government agencies and libraries, are attempting to define roles and responsibilities with respect to the curation of research data. I’ll discuss Cornell University Library’s efforts to: explore and define its own role in this area, partner with other units on the Cornell campus to support data stewardship, and develop an experimental “data staging repository” – a platform meant to promote the sharing of research data sets while research is in progress, and to promote the publication of completed data sets to long-term repositories.