An initiative to create a dataset registry that will support data sharing among researchers from different institutions across the nation and world. Supported by IMLS funding, librarians and IT professionals at Cornell University’s Albert R. Mann Library and Washington University in St Louis are analyzing researchers’ needs and preferences for sharing and discovery of datasets. The DataStaR dataset registry being created will help researchers identify metadata about research datasets in a standard semantic format called Linked Data, and  will be enhanced to support selective interchange of related information with VIVO, an open-source semantic researcher networking tool being adopted by growing numbers of  U.S. universities, federal government agencies, and international institutions.  This work is based on research and results from an earlier, NSF-funded DataStaR project that aimed to create a staging repository for research data.