Pré-publication, Document de travail: Latin America is often portrayed as a global exception to the rising or consolidating income inequality trends of the early twenty-first century. However, the use of administrative data and macroeconomic aggregates casts doubts on this surveybased narrative. In this paper we revisit the region's exceptionalism by building the most comprehensive data base thus far, which accounts for 80% of the region's population and combines harmonised surveys, social security and tax data, and national accounts. We produce a set of inequality indicators-pre and post-tax, based on alternative units and income definitions-which allows us to track the distributional effects of each methodological step and reconcile divergent trends. The reconciliation of micro and macro data present us with a dilemma: either the region is more unequal or it is not as rich as officially reported. The result of distributing the data gaps is a region much more heterogeneous in its inequality trends. Falling inequality is most visible among the bottom 99%, but the trend flattens or reverses in the largest economies once the top 1% and capital incomes are better accounted for. Post-tax and disposable incomes do not change the picture much, except when in-kind social spending is considered. These results confirm the strengths and highlight the limits of Latin America's redistributive policies during the period.