This paper exemines the avilable U.S. data on academic R&D expenditures and the number of papers publiched and the number of citations to these papers as possible measures of "output" of this enterprises.