The purpose of this study was to model the demand for higher education enrollments and to determine whether or not enrollments were price elastic among Virginia's public four-year institutions. The underlying intent was to check the feasibility of a human capital theory based methodology for examining the demand dependency of higher education institutions in Virginia. Price elasticity and other determinants of enrollments were observed across different institutional groupings to observe the differential effects of these factors given institutional type or individual institution characteristics. The analyses were based upon the investment approach to human capital theory as adapted to the study of educational demand. Accordingly, it was assumed that individuals will display a willingness to invest in themselves by enrolling in a college or university because they believe that such an investment will accrue both financial and psychological benefits. Multiple linear regression was used to model five sets of determinants across three levels of analysis: statewide, institutional type (major universities, prior normal schools, and urban institutions), and individual institutions. The direct cost of attendance, the size of the eligible population of prospective students, the educational attainment of the students' locale, and the rural nature of the students' environment generally were found to have significant effects on enrollment. Price, or the direct cost of attendance, was primarily negative and statistically significant for all types of institutions except the major universities and the one special purpose, military institute in the state. Several factors indicated that the location of the institution was important. In other words, students tended to enroll in nearby as opposed to distant institutions, while a geographically concentrated pool of institutions restricted the dependency of these same institutions on local student populations. A cross-sectional design was used in this study, therefore the resultant demand models were descriptive only of the time period covered in these analyses. Nevertheless, such research should be useful in assessing the impact on enrollments of selected demand factors and in determining the efficacy of the investment approach applied in this and similar research efforts. / Ph. D.