In this paper, we examine the hypothesis of market efficiency in euro/dollar with un-anticipated news, which are defined as the difference between actual values and the market's forecasts. The research data are divided into two periods of time, before and after the beginning of financial crisis. Unlike previous literatures in which the un-anticipated news are incomplete and may be unreal, our paper adopted all macroeconomic announcements and indicators of United States and the European Union. Our results before the financial crisis indicate that the market efficiency hypothesis is accepted, although the result fails to hold after the financial crisis. The result still shows the importance of the un-anticipated news in testing the foreign exchange market efficiency hypothesis. Therefore the rejection of efficiency hypothesis on foreign exchange market in the literature may result from the lack of un-anticipated news in the model. In addition, we found that impacts of U.S. and E.U. un-anticipated news are asymmetric on the exchange rate. Besides, the market participants tend to ignore the E.U. news during both periods of time.