There has been much debate concerning the performance of family firms and the drivers of their performance. Some scholars have argued that family management is to blame when family firms go wrong; others claim that family management removes costly agency problems and encourages stewardship. Our thesis is that these disagreements can only be resolved by distinguishing among different types of family firms. We argue that family CEOs will outperform in smaller firms with more concentrated ownership and underperform in larger firms with more dispersed ownership; they will do neither where firms are smaller and ownership is more dispersed or firms are larger and ownership is more concentrated. Copyright © 2012 John Wiley & Sons, Ltd.