Bonuses are used as an inducement to encourage employees to work harder. If a company performs well this year, it will have the ability to grant bonuses to its employees to boost the morale and the competitive drive of the employees. Many companies also give high bonuses to attract and retain talented employees. This effect is termed an incentive effect in previous studies. If the incentive effect dominates, then a positive correlation between bonuses granted and the subsequent company performance should . In other words, more bonuses granted, better subsequent performance. The employee bonus system was regarded in the past as one of the important reasons why high-tech companies in Taiwan could develop rapidly. Because it was viewed as distribution of earnings in accounting and taxed on the par value which is often lower than the market value, the employee bonus system has been widely adopted over the past twenty years and this practice, indeed, has attracted many talented people . However, about eight years ago , foreign investors [of these Taiwan-based companies?] suspected accounting fraud involving the employee stock bonuses, since the bonuses are rewards in reality and therefore should be expensed as salary. Additionally, in order to align with the treatments in other countries, so that the financial statements can be compared among countries, the bonuses would have to be expensed in 2008 and stock bonuses would be taxed on the market value. However, because of the special treatment in accounting and taxation in Taiwan, companies could grant excess bonuses without being reflected on the income statement before the bonuses are required to be expensed. The members of the board, who determine the amount of bonuses given, may also be executives in the company, so they could participate in the bonuses program. When bonuses were perceived as distribution of earnings and did not have to be expensed in the income statement, some companies would grant excess bonuses because the members who determine the bonuses also could get more money and the income number would not be reduced. When making investment decisions, the investors might be misled by the overstated income. Some companies granted bonuses to employees even when the companies did not earn money. In addition, because the stock bonuses were taxed on the par value which is often lower than the market value, the employees who were overpaid just had to pay a little tax to acquire the stocks and to become stockholders. Because the interest of the original stockholders is diluted, the effect is called a dilution effect in the studies. If companies exploit the regulation, we will expect to see a negative correlation between bonuses granted and the subsequent company performance. That is to say, if the bonuses given not because the board members want to motivate employees but because the board members want to take advantage of the regulation, then excess bonuses will be given and the subsequent performance will become worse. The previous literatures in Taiwan which explore the employee bonus system investigate factors influencing the amount of employee bonuses and the relationship between bonus and company performances. The main conclusion is that size, growth opportunity and industry will influence the amount of bonuses given and that an incentive effect exists but if the bonuses were expensed, a dilution effect would dominate. In practice, the bonus policy will be proposed by the board and then voted by the stockholders. So the corporate governance mechanism, like the composition of the board, will influence the bonus policy. Previous studies supported this conclusion. The corporate governance mechanism will influence the execution of the policy and the efficiency of operation so it will affect company performance. Therefore, the motivation of this study is to investigate the relationship between corporate governance and the determination of the bonuses and to explore whether this relationship will influence bonuses to result in an incentive effect or a dilution effect. To investigate the relationship between corporate governance and the bonuses, this study uses regression to distinguish two groups, the excess bonuses group and the adequate bonuses group, and then compares whether there are differences in the corporate governance variables in these two groups. Then I separately investigate the relationship between the bonuses and the performance in these two groups. I want to explore whether companies with bad corporate governance tend to distribute excess bonuses to employees and result in the dilution effect bigger than the incentive effect so the performance become worse next year. However, companies with good corporate governance may not distribute bonuses excessively so the incentive effect is bigger than the dilution effect and the compensation system can reach its goal to improve the company performance. In other words, I claim that whether the bonuses will result in an incentive effect or a dilution effect is related to the corporate governance mechanism. The empirical results support most of the hypotheses. The contribution of this study is that unlike the previous literatures which explore the relationship between the bonuses and the performance using the whole companies and reach the conclusion that bonuses have a positive effect on the performance but they would negatively influence the performance if the bonuses were expensed, this study distinguishes companies with excessive bonuses from companies with adequate bonuses and investigates separately the relationship between bonuses and performance in these two groups. I suppose that bonus itself may be an effective mechanism to encourage employees and its relationship with performance may be influenced by the corporate governance mechanism which will affect the bonuses policy.