In August 2008, Georgia instituted proceedings against the Russian Federation before the International Court of Justice (ICJ) to establish its international responsibility for alleged acts of racial discrimination against the ethnic Georgian population in South Ossetia and Abkhazia by ‘the de facto South Ossetian and Abkhaz separatist authorities […] supported by the Russian Federation’. In order to establish the international responsibility of an outside power for the internationally wrongful conduct of a secessionist entity, it must be shown, inter alia, that the acts or omissions of the secessionist entity are attributable to the outside power. International tribunals usually determine the question of attribution on the basis of whether the authorities of the secessionist entity were ‘controlled’ by the outside power when performing the internationally wrongful conduct. Attribution thus becomes a question of how one defines ‘control’. The test of control of authorities and military forces of secessionist entities has become perhaps the most cited example of the fragmentation of international law. The ICJ, the Appeals Chamber of the International Criminal Tribunal for the Former Yugoslavia, and the European Court of Human Rights have all developed and applied their own tests in order to establish whether a secessionist entity has been ‘controlled’ by an outside power. There is a lot of confusion about the various tests, usually referred to as the ‘effective control’, ‘overall control’ and ‘effective overall control’ tests. This article sets out the various control tests, their requirements and areas of application, and asks which test or tests should be applied to attribute the internationally wrongful conduct of a secessionist entity to an outside power.