In the tournament games only the winner can advance to the next rounds of competition. The important feature of tournament is that the reward is based on the rank order among competitors rather than absolute level of their performances. This type of reward system can be widely found in competition for promotion in business organizations. For example, many hourly workers in manufacturing plants compete for a limited number of supervisory positions; a number of vice presidents in a firm strive for one coveted position of president. The employee participation in improvement activities have been thought of as one of the best practices taken in many auto plants around the world. So we can find auto firms often evaluating their workers on the basis of how good performances they show in off-line activities such as suggestions and QC circles. Or at least it is one of many evaluating items. Since the opportunity of promotions is limited but often known to production workers in advance, it may be regarded as prizes open to candidates. Tournament theory shows that compensation based on relative order can give participants an incentive to provide an optimal level of efforts under certain a set of assumptions. It is often argued that the optimal level of efforts depend positively on the price differential for winning the game. There has been many studies to analyze the benefits and costs that employee participation activities can give individual workers [Cooke (1990)]. However, few attempts have been made to test if promotion tournaments can induce workers to involve more actively in employee participation. In this study I seek to test the incentive effects of promotion using data from a survey of Korean auto plants.