In a university library, variability in circulation of books by subject area is partly but directly dependent on the academic program—e.g. if the anthropology department enrolls more students than the physics department, then more anthropology books should circulate. However, if the number of students enrolled in each subject is the same, the number of books circulated in one subject may still be larger than in the other. Therefore, other sources accounting for variability must be sought. One possible source is in the nature of the subjects themselves. For example, subjects (as represented by academic disciplines or departments) are often called hard or soft, pure or applied, life or nonlife. The purpose of this paper was to determine the relationship between these characteristics and the number of books charged out of an academic library by students. Three major hypotheses were formulated: (1) the softer a subject, the more books charged on that subject; the harder the subject, the fewer the books charged; (2) the purer the subject, the more books charged; the more applied the subject, the fewer the books charged; (3) The more a subject can be characterized as life oriented, the greater the difference in number of books charged between that subject and those characterized as nonlife oriented. A scale value for each characteristic in each of 60 academic subjects was determined by a survey of faculty of the University of Southwestern Louisiana. Correlation and multiple regression were employed to assess the proportion of circulation variance accounted for by each of the three characteristics. Other variables—masters, upper and lower level enrollments, credit hours being taught, number of books already in the library, and level of degree offered—were tested as controls. Hypothesis (1) was weakly supported under the condition that shelflist and masters enrollments were held constant (3% of variance). Support was somewhat firmer (10% variance) when the dependent variable was defined as proportion of shelflist circulated. Hard/soft was also significant (15% of variance) when pure/applied was held constant, but shelflist and enrollments were not. Hypothesis (2) was modestly supported under the condition that other variables were not held constant. Otherwise pure/applied was not significant. Together, pure/applied and hard/soft, in that order, and when no other variables were controlled accounted for 26% of the variance. Hypothesis (3) was not supported under any of the conditions tested. Results have implications in sociological understanding of the relationship between disciplines and library use, in formulating a rationale for library collection building, and in administrative, budget allocations to subjects.