In Chile, reports and research papers have shown that there is an achievement gap in college admissions tests mostly associated to students' gender, socioeconomic status and type of school attended. This gap represents a barrier for low-income and female students to access higher education, as well as for graduates of public schools. Prior studies have used descriptive analyses and single-level linear regression to study this gap, which do not take into account the nested structure of the data (students nested within schools). This study uses multilevel linear modeling to concurrently estimate the effect of student and school characteristics on individual performance in admissions tests in Chile. The findings revealed that more than half of the variation in college admissions test scores happens at the school level. This variation between schools is mostly explained by school sector (private, subsidized private, and public) and the average school socioeconomic status. At the individual level, the most influential factor is individual high school GPA. These findings have important implications for policy and practice, as publicly funded universities in Chile rely almost exclusively on test scores to select students and need-based financial aid requires students to score above a minimum threshold. The results of this study suggest that these admission and financial aid policies need to be reconsidered in order to increase opportunity of access to higher education for traditionally excluded students.