Abstract 
This study aims to investigate the most contributing factors at student-level and school-level in explaining variance in students' achievement in mathematics, using data obtained from Trends in International Mathematics and Science Study (TIMSS) 2019 test, and questionnaires accompanied by it. The study sample included (56,800) students from eighth-grader nested within (209) schools in Saudi Arabia, participated in TIMSS 2019. Hierarchical Linear Modeling (HLM) was used with two (student and school) levels. The results indicated that student and school variables were able to explain (42%) of the total variance. The school level variables accounted for the majority of explained variance (78.59%), compared to student variables which accounted for (21.41%), indicating that the differences in student math achievement is more due to school characteristics. When the variables were controlled at the student and school levels, the statistically significant predictors of students' achievement in mathematics at the student level, were related to potential source of education inequity, including student's background variables related to availability of educational resources at home and the student's characteristics variables related to student's attitudes towards mathematics and school. As for the statistically significant predictors at school level, they included teacher's background variable that relates to his/her specialization in mathematics, and school's climate variables related to percentage of students speaking the language of test, and the school socio-economic status (SES), that is represented by percentage of economically affluent students in the school. Thus, it is important to take these variables into consideration when planning to improve students' education and achievement in mathematics.
Keywords: HLM; TIMSS; Student characteristics; School Characteristics; Mathematics Achievement