International organizations have become key actors in the fight against corruption. Among these organizations, the International Monetary Fund (IMF) maintains a powerful position over borrowing countries in its ability to mandate far-ranging policy reformsso-called "conditionalities"in exchange for access to financial assistance. While IMF pressure can force the implementation of anti-corruption policies, potentially reducing corruption, other IMF policy measures, such as the privatization of state-owned enterprises, can create rent-extraction opportunities and limit the capacity of state institutions to limit corrupt behavior. To test these mechanisms, we conduct instrumental-variable regression analysis using an original dataset on IMF conditionality for up to 141 developing countries from 1982 to 2014. We find that conditions to privatize state-owned enterprises exert significant detrimental effects on corruption control. Conversely, other areas of IMF intervention are not consistently related to corruption abatement. These findings offer policy lessons regarding the design of conditionality, which should avoid large-scale privatization, especially under conditions of weak accountability.