Scientists typically self-organize into teams, matching with others to collaborate in the production of new knowledge. We present the results of a field experiment conducted at Harvard Medical School to understand the extent to which search costs affect matching among scientific collaborators....