We analyze a Bayesian merger game under two-sided asymmetric information about firm types. We show that the standard prediction of the lemons market model-if any, only low-type firms are traded-is likely to be misleading: Merger returns, i.e. the difference between pre- and post-merger profits,...