In this paper we estimate the impact of the statutory labor hours cut in Korea on monthly labor earnings by the regression discontinuity design (RDD) method. The implementation of the statutory labor hours cut policy was sequentially extended based on the number of corporation's employees. The estimation results show that the statutory labor hours cut did not make the workers receiving the treatment better off on average throughout the entire period of 2004-2008 in the sense that it raised monthly labor earnings. However, the policy intervention is found to substantially improve the welfare of workers in the treatment group in 2007 and 2008. Those results imply that the effect of the policy implementation on a worker's welfare depends on the size of the worker's corporation. The policy intervention is more likely to affect workers at large firms favorably than those at small firms. In particular, workers at corporations having at least 300 employees are the main beneficiaries of the statutory labor hours cut policy in Korea.