This dissertation investigates the inequitable distribution and causal impacts of mathematics teachers' content knowledge for teaching (CKT) on instructional quality and student learning outcomes. Using the rich information in the Measures of Effective Teaching (MET) project longitudinal database, the study addresses three core research questions: the systematic inequality of CKT associated with student backgrounds, the causal impacts of CKT on the fine-grained aspects of instructional quality, and the mediating role of instructional quality in the relationship between CKT and student achievement. The research comprises three interconnected studies. The first study uses latent scores of mathematical CKT as a direct measure for teacher knowledge and explores its distribution across different school levels. Unlike conventional qualification indicators such as advanced degrees, CKT is fundamental to instructional quality and significantly associated with students' learning gains. This study employs a novel analytic strategy that accounts for measurement errors in multilevel variance decomposition, providing clarity on the comparative sizes of within-school and between-school variation in teacher CKT. Results revealed that approximately half of the CKT variation lies within schools, while a third is across schools. Notably, although substantial within-school variation exists, it is largely random and not systematically associated with students' prior achievement or socioeconomic backgrounds. However, between-school variations indicate that schools with higher average student achievement and more advantaged socioeconomic backgrounds tend to have teachers with higher CKT. The second study attempts to construct a comprehensive measure of instructional quality that integrates multiple instruments—observational ratings and student perception surveys—to analyze the causal relationship between CKT and instructional quality using a three-level hierarchical linear model. This model accounts for the multilevel clustering of comparable classrooms within randomization blocks at the grade-by-school level, a unique experimental design feature of the MET project. Results found that that higher CKT significantly enhanced the mathematical quality of instruction, particularly in elementary and middle school classrooms. This is reflected in a significant increase in the richness of mathematical content. In contrast to mathematical quality of instruction, no significant CKT impacts were found on other dimensions of instructional quality, such as teacher-student interaction and student perceived classroom experience. With enhanced measurements of CKT and multiple instruments of instructional quality, the third study examines the mediation pathways of instructional quality in the relationship between CKT and student achievement through a multi-step regression approach. This approach allows for a nuanced understanding of how different dimensions of instructional quality might mediate the effect of CKT on student outcomes. While high-school data are restricted by significant reductions in sample size due to missingness in administrative data, the mediation analyses identified noticeable indirect effects of CKT on student achievement through mathematical instructional quality in elementary schools and middle schools. This dissertation provides robust evidence on the importance of CKT in shaping instructional practices and student achievement. By addressing critical gaps and employing rigorous analytical approaches, this research informs future educational policies and practices aimed at enhancing educational equity and effectiveness, ensuring all students receive high-quality instruction from knowledgeable and skilled teachers.