The currencies exchange rates are usually thought to be unpredictable. On the other side, industrial organizations are obliged to hedge their assets by future contracts what requires forecasting the currency exchange rates. In a short run the exchange rates volatility can be predicted using volatility models based on GARCH specification. The paper is aimed to forecast daily exchange rate volatility of such currencies as: the Russian ruble, Indian rupee, Brazilian real, Chinese yuan, Polish zloty and euro against US dollar, using several GARCH specifications with different error distributions. Out-of-sample forecast accuracy was then examined using different measures and procedures. It was to state how much industrial organizations could rely on the results of such forecasts.