The goal of this study was to estimate the relationships among principal leadership, teacher-related factors (e.g. job satisfaction, self-efficacy of teaching, and collaboration), and student achievement in science. With the United States sample of Programme for International Student Assessment (PISA) 2015 data, we used a series of two-level structural equation models to examine the direct, indirect, and moderated effect of principal leadership. The results revealed that (a) teachers' rating of principal leadership had a direct and positive relationship with student science achievement, (b) teachers' job satisfaction and collaboration were positively associated with principal leadership but did not mediate the relationship between principal leadership and student science achievement, and (c) principal leadership's direct effect on science achievement was moderated by school size. Implications are discussed.