Joseph Schumpeter argued that there are increasing returns in R & D both to size of R & D establishment and to firm size. This has been taken to imply that the combination of small firms into big ones would increase R & D output. Tests have been attempted in the literature looking at the relation between scale and R & D input. The present paper shows that these tests are inappropriate. Moreover, the stated conclusion as to R & D output does not follow from Schumpeter's hypothesis. Indeed, it is very nearly true that none of these propositions can be derived from any other.