This study analyzes empirically the effects of product market competition and agency problems on dividend policies of firms listed on Korea Exchange from 2000 to 2013. The sample firms are classified into dividend firms and non-dividend ones, and the dividend firms are classified into chaebol firms and non-chaebol ones. The number of firm-years of dividend firms is 5,008 and that of non-dividend firms is 2,442, and in the dividend firms, the number of firm-years of chaebol firms is 888 and that of non-chaebol firms is 4,120. This study use the industrial concentration measured by Herfindahl-Hirschman index as a proxy variable for product market competition. The main results of this study can be summarized as follows. Firms in more competitive industries are more likely to pay dividends. The positive association between product market competition and dividend payouts suggests that firms in more competitive industries pay more dividends. This result provides evidence that product market competition can be an effective corporate governance mechanism to reduce agency problems between managers and shareholders. Moreover, firms in more competitive industries are more likely to increase dividends and less likely to omit dividends. That is, the increase in product market competition is positively associated with the probability of dividend increases but negatively associated with the probability of dividend omissions. These results provides evidence that the increase in product market competition can effectively force managers to pay dividends to shareholders. Firms with a higher ownership concentration pay fewer dividends, implying that concentrated ownership exacerbates the risk of controlling shareholders expropriating minority shareholders. However, this negative association between ownership concentration and dividend payouts is weaker in industries with intense product market competition among firms. This result provides evidence that product market competition can be an effective corporate governance mechanism to reduce agency problems between controlling and minority shareholders. Low Tobin’q firms with high overinvestment risk are less likely to pay dividends. But low Tobin’q firms with high overinvestment risk in more competitive industries are more likely to pay dividends. This result provides evidence that product market competition can be an effective corporate governance mechanism to reduce agency problems by reducing the high overinvestment risk of low Tobin’q firms. Furthermore, chaebol firms pay fewer dividends relative to non-chaebol firms. This result implies that agency problems between controlling and minority shareholders may actually intensify in chaebol firms, because group affiliation allows controlling shareholders to effectively control the whole firm and to have opportunities to expropriate minority shareholders. However, this negative association between ownership concentration and dividend payouts is weaker in industries with intense product market competition among chaebol firms. This result provides evidence that product market competition among chaebol firms can be an effective corporate governance mechanism to reduce agency problems between controlling and minority shareholders. In conclusion, product market competition that mitigates agency problems between managers and shareholders can also have some effect on agency problems between controlling and minority shareholders. In particular, in Asian capital markets, minority investors tend to be victims of agency problems and have little protection for their investment. Accordingly, this result implies that minority investors need to invest in highly competitive industries with intense product market competition among firms to protect expropriation by controlling shareholders. Therefore, product market competition may be recognized as a new determinant factor which has a significant effect on capital structure in Korean capital market. This paper may have a few limitations because it may be an only early study about the relationship among product market competition, agency problems, and dividend policies of firms listed on Korea Exchange. Therefore, we think that it is necessary to expand sample firms and control variables, and use more elaborate analysis methods in future studies.