We develop a theory to explain the uses and abuses of representative shareholder litigation based on its two most important underlying characteristics: the multiple sources of the legal rights being redressed (creating dynamic opportunities for arbitrage) and the ability of multiple shareholders to seek to represent the collective group in such litigation (creating increased risk of litigation agency costs by those representatives and their attorneys). Placed against the backdrop of controlling managerial agency costs, our theory predicts that: (1) the relative strength of the different forms of shareholder litigation will shift over time, (2) these shifts can result in new avenues for the shareholders to express litigation power, (3) new agents will emerge to act on shareholders’ behalf when these shifts occur (or old agents will put on new hats), and (4) a new set of principal–agent costs resulting from litigation will arise out of these new relationships, leading to recurrent questions about how to best control these costs in particular contexts. Applying our theory to recent academic and practitioner claims of abusive multijurisdictional forum shopping in representative corporate litigation, we conclude that these claims are both overstated and misdirected. Instead, we find a significant amount of what we call “fee distribution litigation.” In these cases, multijurisdictional suits are filed by plaintiffs’ law firms largely to obtain a slice of the total pool of plaintiffs’ attorneys’ fees that are paid in a global settlement in one of these cases. We show that fee distribution litigation is quite different than traditional forum shopping and requires a different policy response. We then consider various approaches and conclude that, while no one of them is perfect, judicial comity is the best and least costly option.