Enrollment in postsecondary education is at an all-time high, and a college education is increasingly seen as a necessity for the majority of students. Prior research has established that several variables account for students' postsecondary enrollment differences, including students' demographic characteristics and their academic achievement. However, despite the wealth of research that shows students' motivation-related beliefs and behaviors are important predictors of academic outcomes and choices, motivation-related theoretical frameworks have rarely been applied to the research examining postsecondary enrollment as an outcome. Therefore, data from the ELS:2002 dataset was used to examine how students' ability beliefs, value for school, postsecondary expectations, and goal-directed behaviors were related longitudinally, as well as how these variables related to students' postsecondary enrollment. Multinomial logistic regression was used to examine whether students' postsecondary expectations changed over time differently for students from different demographic groups. Next, structural equation modeling was used to examine how the full set of motivation-related beliefs and behaviors were related to one another as well as how they predicted postsecondary enrollment. Demographic differences by gender, ethnicity, and SES in the relations between variables were also examined. The results indicate that patterns of change in postsecondary expectations differentially predicted students' postsecondary enrollment. Specifically, students who reduced their expectations were less likely to be continuously enrolled in college. Further, students' ability beliefs and value for school predicted their postsecondary expectations and goal-directed behaviors which in turn predicted students' postsecondary enrollment. Mean-level differences as well as differences in the relations between variables emerged for students' from different demographic groups. Implications of these findings and directions for future research are discussed.