This study examines the applicability of endogenous protection theory, especially “Protection for Sale" model by Grossman and Helpman(1994) for the determination of trade policy in Korea. First, in this study, the basic model based on G-H is developed and examines the effects of labor-related variables on the determination of trade policy by extending the basic model which considers the lobbying by labor unions and labor market rigidities. It is theoretically derived that the labor market rigidity of one sector increase the trade protection level for that sector if the labor suppliers of that sector is politically organized. Second, in this study, the econometric models based on the basic model, the modified basic model with dummy variables for agriculture, and the extended model with labor market rigidities are developed and estimated using the SITC 2-digit level data for Korea. The estimation results do not support the hypotheses of political economy of trade policy. That is, the political economy of trade policy that the trade protection level is high for the sector in which the relatively more gains from protection  and the less cost of protection will be expected because of low import penetration ratio and inelastic import demand is proved not to be working for the determination of trade policy in Korea. The estimation results are partly due to the low level of disaggregation of data. The credibility of estimation results might suffer from the insufficient level of disaggregation of data that might cause high measurement errors. The political game played by industry interest groups and the government in the theory of political economy of trade policy assumes that the only policy considered by players is trade policy(trade protection level in each sector). But, in the real politics, the policies targeted by industry interest groups are not limited to trade policy. Such limitations of the model may result in the above estimation results. However, the estimation results of this study are mainly due to the fact that Korea has reduced trade barriers substantially since 1990s and the differences in tariff rates between industries has been declined due to the policy of equivalent tariff structure. The other important characteristic of the estimation results is that whether the sector belongs to agriculture is very critical to the trade protection level of that sector. This characteristic shows that the determination process of trade policy in Korea is very influenced by populism. The policy-makers are very likely to cater to the sentiment of median-voters who prefer to the protection of agriculture but less likely to respond to the political lobby of industry interest groups. The estimation results of this study have important implications for the strategy that minimize the internal political opposition to trade liberalization such as trade adjustment assistance program. Since, from empirical results of this study, it is rarely found the evidence that industry interest groups are effectively lobbying the government for trade protection in Korea, the program that promises to help the sectors suffering from trade liberalization will not be able to alleviate the political opposition to trade liberalization. on the contrary, for sustaining more trade liberalization, it is imperative to overcome the influence of opinions of opposition groups on the sentiment of median-voters.