The present study addresses the question of whether school leadership matters. It employs second-order meta-analysis to synthesize results from 12 first-order meta-analyses examining school leadership effects published 2003-2019. These meta-analyses collectively examined 512 primary studies published across four decades (1978-2019). Results showed that the overall mean effect size for school leadership was small in magnitude (r = 0.33). Effect sizes for leadership models were larger than those for leadership practices, thereby indicating the utility of examining models as compared to practices for understanding leadership influence. Relatedly, findings of significant positive effects for eight different school leadership practices underscore the need to examine comprehensively the scope of school leaders' work beyond that related to teaching-and-learning. Additionally, leaders require myriad competencies and skills including how to galvanize, motivate and equip teachers to achieve school goals. The substantially larger mean effect sizes for organizational and teacher as compared to student outcomes challenge the assertion by some that principals are less consequential than teachers in contributing to school effectiveness. Indeed, the larger effect sizes for principals as compared to other types of leaders reflect the key role they play in leading schools.