This study revisits the market reactions on the dividend payment events, cum-dividend date and payment date, using the event study method. The sample of this study includes all dividend announcements from 2017 to 2018 in the Indonesia Stock Exchange. This study performs various robust statistical tests proposed by Harrington and Shrider (2007), who point out that standard classical t-test is not enough to ensure abnormal return on an event because of the bias due to volatility caused by an event. Using various statistical tests for testing the abnormal return, this study shows that the market indeed reacts to the cum-dividend date and dividend payment date, as well as showing that the classical t-test showing the same conclusion as the other tests.