This study investigates the nonlinear relationship between gender inequality and gender gaps in math achievement across 86 countries using Programme for International Student Assessment data. While previous research has assumed a linear relationship, this study finds U-shaped associations between the Gender Gap Index (GGI), and some of its components and gender gaps in math scores. Specifically, initial improvements in gender equality seemingly widen the gap, while further progress narrows it when boys outperform girls, although the patterns differ when girls outperform boys. Upon excluding Muslim-majority countries, where girls often outperform boys despite the high gender inequality, many relationships shift from nonlinear to linear. Additionally, this study highlights the limitation of relying solely on the GGI that often fails to reflect female advantages in education consequently yielding inconclusive results. The findings underscore the importance of disaggregated analysis and caution when interpreting prior studies that overlook nonlinear patterns and sample composition.