Empirical investigations on corruption have used surveys on the perceived level of corruption in a cross-section of countries. The chapter explains the methodology of the Transparency International Corruption Perceptions Index 1999 and presents the various contributing sources. The article claims that combining the sources into a composite index is a valid means of improving the robustness of the survey results. The high correlation between various sources is an indicator that subjective assessments improve the understanding of real levels of corruption. The Index included data from nine sources. Other sources were considered and rejected because they mixed corruption with other issues such as xenophobia and nationalism, or concerned changing corruption rather than levels of corruption. Ninety-nine countries were included in the Index. In explaining the methodology of the Index, the article includes discussion of objective vs subjective data; year-to-year comparisons; terminology definitions; reliability; interpreting perceptions; the role of the media; weighting the data; and standardizing. Notes, tables, references