Using a global sample of 26,444 M&A transactions, we empirically iestigate the main drivers of serial acquirer takeover performance, focusing on the benefits of organizational learning and the detriments of post-merger integration (PMI) problems. Taking single acquirers as a control group, we find that multiple deal makers exhibit a 0.4 percentage points lower short-term abnormal return, equal to a $31M shareholder value disadvantage. While learning through experience does not depend on the quantity of acquisitions, serial acquirers can gain proficiency in some specific types of deals (specialized learning hypothesis). When looking at the PMI process, the time between two consecutive transactions and relative deal size are key success factors, suggesting that the capacity of an organization to integrate is limited. This is strong support for the existence of the indigestion hypothesis. Furthermore, we are able to document that overconfidence is a driving force of such problems.