Do alternative exchange rate regimes affect short-term real exchange rate volatility differently? The existing empirical evidence is quite mixed with slightly more papers supporting that they do. We show that such lack of consensus is mainly due to current literature limitations regarding the measurement of real exchange rates (RERs), the identification of exchange rate regimes (ERRs), and the control for the incidence of real and nominal shocks. To address these limitations, we construct a novel monthly dataset for 63 countries over the period 1946–2007, which includes market-determined multilateral RER and a proxy for terms of trade. We find that ERRs indeed affect short-term real exchange rate volatility differently. While the evidence is generally consistent with Mussa's sticky prices argument, we find that for nonadvanced countries in post-Bretton Woods there exists a “U-shape nominal flexibility puzzle of RER.” We also find evidence of a “short-run RER volatility puzzle.” Having controlled for the incidence of real and nominal shocks, nonadvanced countries' RER volatility remains between 25% and 150% greater than that of the advanced economies. Moreover, the key literature finding that short-term RER volatility is higher in Bretton Woods (BW) than in post-Bretton Woods (PBW) for industrialized countries vanishes when using market-determined multilateral RER instead of official bilateral RER. (JEL F31, F33, F41)