<p>This study examines how board gender diversity and board size shape the profitability of commercial banks in Ethiopia. As the financial sector undergoes transformation, especially in emerging economies, understanding these governance dynamics is increasingly relevant. Drawing on a balanced panel of 17 Ethiopian banks from 2010 to 2021, we analyze the impact of corporate governance mechanisms using Return on Assets (ROA) as the key performance metric. Fixed Effects models form the core estimation strategy, selected through the Hausman test, and supplemented by Panel-Corrected Standard Errors and Generalized Method of Moments to check robustness and address endogeneity concerns. The results offer a nuanced picture. Larger boards are consistently associated with improved profitability, defying traditional views that caution against board expansion. In contrast, board gender diversity shows no clear direct effect, but it becomes significant in interaction with board size—though the moderating effect is negative. This suggests that while diversity brings value, it may also dilute the benefits of larger boards if not backed by structural support. Bank size and liquidity generally enhance performance, while customer deposits unexpectedly show a negative association in some models. The effect of economic growth is inconsistent—negative in FE models but positive when endogeneity is corrected via GMM—while inflation appears largely irrelevant across specifications.These findings point to the context-specific nature of governance in developing banking systems. They challenge assumptions embedded in agency theory and suggest that governance reforms need to move beyond surface-level representation. For Ethiopian regulators and policymakers, the lesson is clear: tailoring governance policies to local realities may prove more effective than importing global best practices wholesale.</p> Graphical abstract <p></p>