In the banking industry, the structure-performance relationship has frequently been evaluated with results suggesting that collusive profits occur. These studies have been criticized for inappropriately accounting for entry barriers, ad hoc assumptions concerning the appropriate structure measure, limited samples, and ignoring firm efficiency differences. We address these concerns and find categorical support for the efficient structure hypothesis, and limited support for the traditional structure-collusion hypothesis when markets are characterized by significant entry barriers. The findings suggest the competing hypotheses may actually be complementary theories, and the negative role of entry barriers may be more important than previously thought.