We utilized four waves of TIMSS data in addition to the information we have collected on countries' educational systems to examine whether different degrees of standardisation, differentiation, proportion of students in private schools and governmental spending on education influence students' math achievement, its variation and socioeconomic status (SES) gaps in math achievement. A higher level of standardisation of educational systems was associated with higher average math achievement. Greater expenditure on education (as a percentage of total government expenditure) was associated with a lower level of dispersion of math achievement and smaller SES gaps in math achievement. Wealthier countries exhibited higher average math achievement and a narrower variation. Higher income inequality (measured by the Gini index) was associated with a lower average math achievement and larger SES gaps. Further, we found that a higher level of standardisation alleviates the negative effects of differentiation in the systems with more rigid tracking.