Books represent a mass medium of communication which has received comparatively little attention from students of public opinion. In this article an attempt is made to evaluate some of the effects of the recent mushrooming of book clubs on the publishing of trade books as a whole, and to this end an experimental method of comparing the merit of various classes of books has been devised. The results of this comparison may surprise both friends and foes of the book clubs: in general, it is found that the clubs do not lower literary standards, and may raise them in some cases, although much depends upon the type of public which each book club is attempting to reach. The author is associated with the Bureau of Applied Social Research, Columbia University.