Chapter 3 outlines and tests our theory of IO exit by applying it to the predictors of IO withdrawal. We argue that many dissatisfied states use the process of withdrawal to broker deals for institutional change in the IO. Many withdrawals are driven by preference divergence from other member states or declining power. Using our IO Exit dataset, we analyze 387 IO withdrawals from 1913 to 2022 across 534 IOs and 198 states. In categorizing the reasons for state withdrawals, we show that two-thirds of IO withdrawals are motivated by the desire to negotiate change rather than by issues that reflect populism, nationalism, or capitulation toward international cooperation. States also use the threat of withdrawal, which supports the notion that exit is a negotiating process with multiple steps rather than a final or singular act. Withdrawal is usually not permanent; half of the time, states return to the IOs they left. States also likely consider costs a priori and avoid withdrawal if the costs are projected to be too high. This prevents many withdrawals from happening in the first place. We do not find consistent support for alternative arguments that backlash against globalization, encroachment from authoritative IOs, nationalism/populism, or legal rules are robust drivers of withdrawal.