SUMMARY We test the predictions of recent theoretical studies of the impact of sanctions on the exchange rate. We build a database of exchange rates and sanctions spanning 1914–45 – an era when both large and small economies were targeted by multilateral sanction packages, facilitating comparisons with today’s Russian war episode. We estimate the dynamic response of the exchange rate in a panel of sanction episodes at weekly frequency using local projections, conditioning on the type of sanctions taken. We tease out mechanisms through which sanctions affect the exchange rate by estimating their effects on macroeconomic variables plausibly acting as transmission channels. Our estimates suggest that import restrictions, export restrictions, asset freezes and trade embargoes lead to exchange rate effects consistent with theory, though the precision of the measured effects varies across sanction type. These findings suggest that recent models of the effects of sanctions on the exchange rate do not just match developments in today’s specific Russia episode but have broader applicability. It follows that the direction of exchange rate movements is not an adequate metric of the success or failure of sanctions but a reflection of the type and scale of the measures taken.