The common treatment of a separate import demand specification in the literature is usually motivated by product differentiation. Most studies, however, proceed further with a separability assumption between domestic and imported product for data consideration and ease in estimation. In this study, a two-stage model is used to estimate aggregate and source-specific import demand elasticities for pork in Japan. This approach allows substitution between domestic and imported product on the one hand and avoids econometric problems in generating source-specific parameters, on the other. Pork imports into Japan are constrained by both the high protection and the strong preference of Japanese consumers for domestic pork over imported pork. Domestic pork commands a price premium of 13 to 29 percent in the retail market. Also, imported pork has a relatively low income elasticity reflecting consumer survey results of lower quality rating for imported pork compared with domestic pork. U.S. pork exports to Japan in particular have lower income elasticity than their closest competitor--Canadian pork. Japanese consumers perceive U.S. pork as inexpensive, but with food safety and quality being the main drivers of pork import demand in Japan, an "inexpensive" attribute may not be the right signal that will provide a true market advantage. However, U.S. pork exports to Japan have performed very well in the last three years. This means either that the United States was just strategically positioned when other foreign suppliers such as Taiwan and the European Union (EU) were challenged by diseases (e.g., foot-and-mouth disease [FMD] and Classical Swine Fever [CSF]), or that U.S. exporters are getting better at understanding Japanese consumer preferences, and are delivering the products that meet them.