This research aims at examining dividend-signaling hypothesis by testing the relationship between dividend changes and operational performance as well as market performance. The sample consists of 75 firms for the period 2000-2005. The analysis employed are multiple regressions and event study. The results on operational performance show that (1) dividend increase (with earnings increase or earnings decrease) correlates significantly with the increase of next period company 's profitability, (2) dividend decrease and earnings decrease negatively correlate with next period company's profitability, (3) dividend decrease and earnings increase positively correlate with next period company 's profitability, and (4) unchanged dividend does not correlate with next period company's profitability. The evidence on market performance (cumulative abnormal return surrounding dividend announcement date) shows that dividend changes have significant effect on cumulative abnormal return for five and ten days around announcement date. Evidences from this research show that dividend signaling hypothesis is not fully supported.