The annual Corruption Perceptions Index (CPI), published by Transparency International (TI), has had a pivotal role in focusing attention on corruption. Despite recent critiques of the CPI, it remains highly influential on research into the causes of corruption and is also extensively used to galvanise support for measures to fight corruption. In this article we explore the CPI in more depth in order to highlight how the index has been used for political ends which may not always turn out to be supportive of anti-corruption efforts. The argument is developed in four sections: in the first, we focus on Transparency International's definition of corruption, highlighting some conceptual difficulties with the approach adopted and its relationship to the promotion of ‘good governance’ as the principal means of combating corruption. In the second section, we outline some methodological difficulties in the design of the Corruption Perceptions Index. Although the CPI has been much criticised, we demonstrate in the third section that the index continues to exercise great influence both in academic research and in the politics of anti-corruption efforts, particularly as exercised by Transparency International itself. In the final section we argue that the CPI contributes to the risk of creating a ‘corruption trap’ in countries where corruption is deeply embedded, as development aid is increasingly made conditional on the implementation of reforms which are impossible to achieve without that aid.