To cope with the recent financial crisis in Southeastern Asia, the government ofTaiwan launched a series of economic reforms. Among them, the Bank Business Tax Reduction Act (hereafter, the Act) reduced the bank business tax rate from 5% to 2%. The whole enactment process took five months with five major events. This study adopts a multiperiod event approach to empirically investigate the impacts of the Act. And three methods (portfolio test statistic, standardized test statistic, and dummy variables in the market model approach) are used to test the significance of the abnormal returns in order to examine the robustness of the results. Also, the U statistic (Patell, 1976) is employed to test the volatility of stock returns in some situations. The findings show that the Act impacted the security returns on four of the five event dates with significantly positive abnormal returns. The existence of significantly positiv_ecumulative two-day and five-day abnormal returns before the announcement of the Act by the Ministry of Finance (1999/2/20) and the approval by the Executive Yuan to the Act (6/3) implies possible information leaks or information searches by investors. The post-event drifts after 2/20 and 3/2 (the late President Lee expressed an opinion showing that the bank business tax rate should be reduced to zero) suggest that investors took time in the early part of the enactment process to figure out the full impacts of the Act and the probability of its fmal approval, or the results may be due to the regulated 7% market fluctuations. The only situation that can be tested for the volatility of stock returns is the cumulative two-day abnormal return after 6/17 (the date that the Finance Committee of the Legislative Yuan approved the Act). Even though its abnormal return is zero, its U statistic is significant. It shows that the variance of stock returns has been significantly increased.