The primary purpose of this study was to construct and test a model of the factors affecting high school students' first time entry into different types of postsecondary institutions. To test the model, a sample (N = 5395) of high school students was drawn from the sophomore cohort of High School and Beyond data. These students were enrolled in different types of postsecondary institutions: four year colleges and universities, two year junior colleges, community colleges, and vocational technical colleges. The path model is a set of structural equations that considers the college enrollment decision of the students to be a function of four exogenous variables and six endogenous variables. The exogenous variables were: socioeconomic status of the student's family, student's academic ability, academic performance, and educational aspiration. The endogenous variables in the model were: high school program, encouragement for college attendance, cost of the postsecondary institution, size of the postsecondary institution, and student aid both in the form of grant and loan. These variables were arranged in a fully-identified block recursive model. Because of possible interactions caused by different parameters between blacks and whites, the model was analyzed separately for black students and white students. The model also was estimated separately for male and female students. The computer program, GEMINI, was used to estimate the model. Results indicated that tuition cost, academic ability, and educational aspiration were the most important influences on students' enrollment choice between four year colleges and other postsecondary institutions. Financial aid variables, both grants and loans, exerted significant positive effect on the college enrollment decision as well. The effects of these variables were found to be similar for blacks and whites, and for males and females. Recommendations for future research include further work on college going behavior with different populations. The studies of non-traditional patterns of attendance and the impact of current financial aid policy on these patterns would contribute to better understanding of college attendance behavior. / Ph. D.