<p>Research consistently indicates that metacognition is associated with academic achievement across various content areas, particularly in mathematics. However, limited research has explored how students across different proficiency levels utilize metacognitive skills and their impact on mathematical performance. This study investigated the performance levels in a standardized math test and the application of metacognitive skills (planning, management, monitoring, evaluation, and reflection) through a questionnaire among 46,346 eighth-grade students from 128 secondary schools in China. Employing the Angoff method, participants were classified into four distinct levels of math achievement (A, B, C, D levels). Students with a C level or higher demonstrated compliance with the stipulated requirements of the standard. The results showed there were significant differences in metacognitive skills among these levels. Findings also revealed a positive correlation between metacognitive skills and mathematics achievement (<i>r</i> = .299, <i>p</i> &lt; .01), with planning demonstrating the strongest association with math performance and reflection showing the weakest. Moreover, students with higher math proficiency exhibited increased application of metacognitive skills, highlighting the relationship between self-regulation in problem-solving and academic success. These findings emphasize the relationship between metacognitive skills and mathematical performance, proposing targeted educational interventions informed by the application of metacognitive skills across different proficiency levels.</p>