A case is made for targeting real exchange rates under the extended target zone proposal. Wide, rather than narrow, bands are preferred to allow the real exchange rate to adjust to shocks via changes in the nominal exchange rate rather than in the price level. Relatedly, wide bands allow monetary control to anchor the price level. Wide bands are cost-efficient as they do not attempt to eliminate real exchange rate volatility that is of little economic consequence. Credible wide bands will also help to dampen trends caused by technical-driven speculation. Empirical evidence relevant to the main arguments is reviewed.