This paper reconsiders some conventional notions about fiscal policy under flexible exchange rates using an extended version of the well-known Dornbusch ',overshooting" model. Three widely held views are challenged. These are: (1) the Mundell-Fleming result that fiscal policy is ineffective under flexible exchange rates; (2) that real shocks (including fiscal policy) do not cause exchange rate overshooting; and (3) that real shocks are not important in explaining high exchange rate variability. In our model output is affected by fiscal policy both in the short run and long run". It is possible for unanticipated fiscal policy to cause perverse output effects in the short run; anticipated future fiscal policy necessarily leads to a fall in current output. Our results also show that fiscal policy can lead to exchange rate overshooting even in the case where monetary policy also does. Further, in this model both fiscal and monetary policy must cause either nominal exchange rate overshooting or perverse domestic interest rate effects. Finally, the results suggest that real shocks are a potential source of exchange rate variability.