Chapter 6 details and tests our theory of IO exit by applying it to the predictors of IO suspensions. IO member states use suspension to punish states that have violated IO commitments and to incentivize domestic institutional change. We argue that suspension is not an automatic punishment for violations but instead is influenced by factors related to bargaining and institutional constraints: Violator states that are more powerful, have material resources, and have alliance relationships with regional powers are less likely to be suspended while IOs. Empirically, we analyze 101 IO suspensions from 1939 to 2022 across all IOs and states, and then focus our multivariate analyses on suspensions for political backsliding. This is because we show that most suspensions occur for human rights violations and incursions on democracy commitments (like coups d’état); and narrowing the scope allows us to control for the kinds of violations that prompt suspension. Our quantitative analysis shows that IO membership suspension is imposed against some but not all violators – and that this is partly because powerful states are able to insulate themselves from IO pressures, avoiding punishment for violations that less powerful states get suspended for. IO institutional constraints including their democratic density also affect the likelihood of suspension for political backsliding. Suspension can act as a multilateral diplomatic sanction but power and politics matter.