Abstract: This paper empirically documents the causes and consequences of mergers and acquisitions in the US video game industry. For this purpose, we use a widely used data set from NPD on video game monthly sales from October 2000 to October 2007 that we complement with hand-collected information on the identity of video game developers for all games in the sample and the timing of all mergers and acquisitions during that period. During this period of time, we observe that 57 out of roughly 600 upstream game developers were acquired by downstream publishers. Using Whinston (2003) as point of departure, we build up a toy model of vertical integration in the presence of non-contractible specific investments and derive testable implications that we take to our data. Then, we extend the model to a dynamic setting where developers and publishers learn about the quality of the match between them. Our preliminary findings show that (1) a developer is more likely to get acquired after releasing a hit game, (2) publishers and developers are likely to underperform after an acquisition takes place, (3) the acquiring party may not need to have contracted in the past with the acquired party, and (4) the impact of developer and publisher location and game portfolio on acquisition decisions play an important role to determine the goodness of the match between developer and publisher. While we argue that these findings are difficult to reconcile with Transaction Cost Economics theories, theories of integration that focus on the relevance of non-contractible investments, learning and employment relations do not seem to do much better.