This study utilizes a linear mixed-effects regression model to examine how regional differences in education systems across Brazil impact student mathematics achievement. Employing a Bayesian multilevel framework, it analyzes region-level, school-level, and student-level variables while considering the hierarchical nature of the data. A comparative analysis of PISA scores, conducted before and after the BNCC implementation in Brazilian schools, evaluates mathematics achievement. The findings indicate that both regional factors and school type significantly affect variations in achievement. Notably, a performance gap persists between the Northern and Southern/Southeastern regions, reflecting disparities in access to resources. Furthermore, the differences between public and private schools highlight systemic inequalities in educational opportunities. Overall, both student-level and school-level factors are crucial in shaping students' mathematical achievement, as evidenced by the PISA scores.