The exchange of data sets within the operations research community suffers from three main issues: (1) data sets are exchanged as plain text files; (2) data sets are offered on centrally managed websites; and (3) the results of applying algorithms to these data sets are unavailable. These issues result in an inefficient sharing of research artifacts. In this paper, we present libOR, a Web-based library of data sets for the operations research community. The organization of libOR is based on the open source and open content approach. The project has three main objectives: (1) stimulate information sharing of research data; (2) increase collaboration to increase scientific advancement; and (3) stimulate learning from approaches undertaken by other researchers in the domain. Early feedback from operations researchers seems to indicate that the advantages offered by libOR are greatly welcomed.