The study of matching problems typically assumes that agents precisely know their preferences over the goods to be assigned. Within applied contexts, this assumption stands out as particularly counterfactual. Parents typically do invest a large amount of time and resources to find out which school would be best for their children, doctors run costly tests to establish which kidney might be best for a given patient. In this paper I introduce the assumption of endogenous information acquisition into otherwise standard house allocation problems. I find that there is a unique ex ante Pareto-optimal, strategy-proof and non-bossy allocation mechanism: serial dictatorship. This stands in sharp contrast to the very large set of such mechanisms for house allocation problems without endogenous information acquisition.