Although enrollment in advanced mathematics courses is a significant determinant of mathematics achievement, the majority of public school students are not enrolled in advanced mathematics courses in high school. Policy makers are interested in the dynamics of the math course selection process in the eighth grade because it is viewed as a pivotal transitional point when students are confronted with the decision to either enroll in algebra, the first course on the advanced math track, or in regular math. Approximately one third of eighth grade students enroll in algebra, in spite of general availability of the course. Enrollment patterns vary among the four major race/ethnic subgroups - Asian, Hispanic, Black and White. This study constructed and tested a structural equations model that examined the factors influencing math course choice and the course selection process in the eighth grade in public schools. There were three sources of influence in the model: 1) math achievement; 2) school policies and practices; and 3) parents. The model consisted of three exogenous and five endogenous variables. The model was tested five times. It was tested on a nationally representative sample of 7,648 eighth grade public school students. It was also tested separately on the four race/ethnic subgroups comprising the full sample. The study used data from student and parent files of the base year survey of the National Education Longitudinal Study of 1988 (NELS 88), a major national study conducted under the auspices of the National Center for Education Statistics (NCES). For the full sample, the major school and parental factors influencing a student’s math course choice were math track placement, parents’ educational expectations and school-parent algebra push. Of the two achievement influences, standardized math test scores had the stronger influence on the outcome variable. Prior math grades influenced math course choice, but to a lesser extent and was influential largely due to an indirect effect. Although these factors were important for each of the subgroups, the influence of the factors varied among the subgroups. The model fitted the data fairly well for the full sample and the Asian and White subgroups, but less well for the Hispanic and Black subgroups. / Ph. D.