This paper presents a reclassification of historical exchange rate regimes based on market-determined parallel exchange rates, revealing significant differences from the official classification. The authors find that roughly half the time, official pegs actually represent different monetary regimes, often floats. Conversely, when official classifications are floating, the reality is often a de facto peg. The new classification shows that the breakup of Bretton Woods had less impact on most exchange rate regimes than commonly believed. The authors also find that exchange rate arrangements may be important for growth, trade, and inflation. Their monthly data set on market-determined exchange rates goes back to 1946 for 153 countries. The paper argues that the official classification is flawed, as it often fails to reflect actual country practices. The authors propose a "natural" classification system that uses historical chronologies and market-determined exchange rates to better classify exchange rate regimes. They find that dual or multiple rates were more common than previously thought, and that de facto floating was not uncommon during the Bretton Woods era. The authors also find that freely falling regimes, associated with high inflation, are not uncommon. The paper concludes that the official classification is inadequate and that a more accurate classification is needed to understand the impact of exchange rate regimes on economic performance.This paper presents a reclassification of historical exchange rate regimes based on market-determined parallel exchange rates, revealing significant differences from the official classification. The authors find that roughly half the time, official pegs actually represent different monetary regimes, often floats. Conversely, when official classifications are floating, the reality is often a de facto peg. The new classification shows that the breakup of Bretton Woods had less impact on most exchange rate regimes than commonly believed. The authors also find that exchange rate arrangements may be important for growth, trade, and inflation. Their monthly data set on market-determined exchange rates goes back to 1946 for 153 countries. The paper argues that the official classification is flawed, as it often fails to reflect actual country practices. The authors propose a "natural" classification system that uses historical chronologies and market-determined exchange rates to better classify exchange rate regimes. They find that dual or multiple rates were more common than previously thought, and that de facto floating was not uncommon during the Bretton Woods era. The authors also find that freely falling regimes, associated with high inflation, are not uncommon. The paper concludes that the official classification is inadequate and that a more accurate classification is needed to understand the impact of exchange rate regimes on economic performance.