Trends in International Mathematics and Science Study (TIMSS) has been taken place for almost 20 years in order to determine the trends in mathematics and science achievement of fourth and eighth grades students. This means that cross-sectional data can be obtained by TIMSS so as to see which variables play role in students’ achievement at different grades in each administration. This study examined student and teacher level variability of the students’ mathematics achievement for each grade in terms of TIMSS 2011 Turkish data. The common variables adressed in both 4th and 8th grade students' questionnaire and their teachers' questionnaire were selected to be compared. More precisely, confidence in teaching Mathematics, teacher working conditions, school emphasis on academic success, teacher career satisfaction, collabarate to improving teaching, and instruction to engage students in learning were selected as teacher level variables addressed in teacher questionnaire.  As student level where addressed in student questionnaire were  being bullied at school, students’ engagement in Mathematics, students’ confident in mathematics, students’ like learning Mathematics, belogning to school, and parents’ involvement in students learning. Due to the nested structure of the data in TIMSS, hierarchical linear modeling analysis was performed using HLM 7 software. According to the findings of the study, for both grade level, of all teacher level  variables, only the variable school emphasis on academic success perceived by teachers had statistically significant effect on schools' mean mathematics achievements. Moreover, of all the student level variables,  being bullied at school, students’ confident in mathematics, students’ engagement in mathematics and parents’ involvement in students learning variables were found to have statistically significant effects on students' mathematics achievement for both grade level. Findings from this study will provide a potentially fruitful contribution to related literature.