Doctor of Philosophy / Department of Agricultural Economics / Tian Xia / This dissertation presents the development of a new approach to include the interaction of vertically differentiated products, a subject that has been largely ignored in previous studies, to analyze the market power of exporters and importers in the world markets of agricultural commodities. Three theoretical models, a residual demand elasticity (RDE) model, a residual supply elasticity (RSE) model, and a two-country partial equilibrium trade model, are developed, and the corresponding empirical models are specified for U.S.-Japan soybean trade. Genetically modified (GM) and non-genetically modified (non-GM) soybeans are vertically differentiated products in the sense that GM soybeans are largely defined as an inferior substitute to non-GM soybeans. I compare two versions of these models: a new approach in which the interaction between non-GM and GM soybeans is taken into account and the traditional approach in which the interaction is ignored. In each of the three models (the RDE model, the RSE model, and the partial equilibrium trade model), the traditional approach overestimates the market margin of U.S. non-GM soybean exporters and that of Japanese non-GM soybean importers. By considering the interaction between non-GM and GM soybeans, the new approach greatly reduces the estimates of the corresponding market margins of U.S. exporters and Japanese importers to improve the accuracy of such estimates. The statistical significance of the coefficient estimate of the interaction term, the U.S. GM soybean price or the Japanese GM soybean price, in all three models suggests that the new approach, which includes the interaction between non-GM and GM soybeans, is necessary and preferred. The partial equilibrium trade model includes both an RDE equation and an RSE equation in a system to address the possible contemporaneous cross-equation correlation. Thus, the estimation results of the partial equilibrium trade model are further improved, compared to those of the RDE model and the RSE model. Using the traditional approach to estimate the partial equilibrium trade model, I find that the U.S. non-GM soybean exporters’ market margin is 56.5% and the Japanese non-GM soybean importers’ market margin is 16.1%. However, the results obtained by using the new approach show that the market margins of U.S. exporters and Japanese importers are 33.2% and 6%, respectively. By taking into account the interaction between non-GM and GM soybeans, the new approach improves the accuracy of the estimates of market margins of soybean exporters and importers. U.S. non-GM soybean exporters do have a significant market margin in international markets, but it is not as large as the one suggested by the traditional approach. Although Japanese non-GM soybean importers enjoy some market margin, it is relatively small. The theoretical and empirical models and results in this dissertation provide new and more accurate estimates of residual demand and supply elasticities and market power and improve the understanding on world soybean markets. These results can be useful for industry participants in international soybean markets, academic researchers, and policy makers.