This paper provides a comprehensive analysis of the disagreements between the two most popular but also discordant de facto exchange rate regime classifications: the Reinhart and Rogoff and the Levy-Yeyati and Sturzenegger classifications. We estimate probabilities of disagreement between the two classifications for the different exchange rate regime categories, and derive a de facto synthesis classification, using the Receiver Operating Characteristic analysis. We show that more than a third of the observations are not directly comparable, and relatively few disagreements are directly attributable to the classifications’ key variables. Most of the disagreements originate from the different thresholds used by the classifications in the definition of the ERR categories and the interactions between several variables. Given these complexities, the synthesis classification provides a useful framework in terms of greater comparability.