Purpose: Greater understanding about how variables mediate the relationship between leadership and achievement is essential to the success of reform efforts that hold leaders accountable for student learning. The purpose of this paper is to test a model of integrated transformational leadership including three important school mediators. Design/methodology/approach: In total, 5,392 students provided data on student engagement, and 569 teachers provided survey data on principal leadership and the collective teacher efficacy (CTE) and extra effort of their colleagues. Data were analyzed at the school level (n = 51) using mediation analysis with the PROCESS macro to calculate the direct and indirect effects of all of the variables in the serial multiple mediator model. Findings: The model explained a significant amount of variance in both reading (38 percent) and mathematics (35 percent). CTE and student engagement mediated all of the leadership effects, demonstrating the importance of teacher and student relations in school improvement. The combined effect sizes are significantly greater than the effects of earlier TL models. Originality/value: This study contributes to "leadership for learning" research by demonstrating that refining transformational leadership behaviors to include instructional management can enhance a leader's impact on achievement. As one of the few studies to examine student engagement as a mediator along with teacher mediators, the current study highlights the importance of teacher-student relations in school improvement. While collective efficacy mediated all of the effects of leadership on achievement, student emotional engagement also contributed to mathematics achievement. The study's findings advance research into mutual influence processes, indicating that reciprocity may extend throughout the leader-teacher-student loop.