Our study demonstrates empirically that the choice of resource allocation strategy affects innovation performance. Allocating resources to a broader range of innovation projects increases new product sales, an effect that appears to outweigh that of resource intensity. In addition, we find that the performance benefit of breadth is higher for firms that allocate resources selectively at later stages of the innovation process. This breadth-selectiveness effect is greatest for firms intending to create relatively more novel products, departing further from their knowledge base. Based on these results, we theorize that breadth increases performance because it spreads firms' bets on unproven innovative endeavors. Limiting resource commitments by selecting out deteriorating projects prevents an escalation in the costs of breadth. This advantage increases with the uncertainty implicit in greater innovative intent. The paper thus contributes to theory of how resource allocation strategies influence performance outcomes of innovation project portfolios. Copyright © 2013 John Wiley & Sons, Ltd.