A general equilibrium model with heterogeneous agents is presented to explain why the undeveloped economies are not industrialized until the late twentieth century. The model takes into account two factors related to the geographic structure of the economy. First, the poor infrastructure of ah undeveloped economy leaves its domestic market spatially segregated because of the high transport costs. Second, depending on how readily the labor force in the rural areas can move into the industrial area, the cost of labor may be tao expensive to run a factory profitably. The model bears several policy implications. The government`s investment in infrastructure encourages the industrialization. The effect of the government`s investment in the infrastructure on the geographic concentration of industries depends on the fixed cost of the manufacturing technology. A country ran get around the problem of high transportation costs by targeting foreign markets and exploiting the cheaper ocean transportation costs. A reduction in the cost of migration, such as housing costs and the cost of risk-sharing, the labor force with low reservation wage can be attracted to the industrial area. Although the industrialization has spread out all over the world throughout this century, many countries still remain undeveloped until now. (Easterlin, 1981) Explaining why it is so difficult for some economies to launch the industrialization, the model in this paper takes into account two major factors related to the spatial structure of an economy. First, the poor infrastructure of an undeveloped economy leaves its domestic markets spatially segregated. In such I thank Robert Townsend, Nancy Stokey, and D. Gale Johnson for the comments. I also appreciate Joon-mo Yang`s comment at the annum conference of Korean Economic Association in 1998. I also thank the anonymous referee for the valuable comments. This research was supported by Yonsei University Research Grant (1998). circumstances, each of the localized and fragmented domestic markets of an undeveloped economy may be too thin to cover the fixed costs of the modern industrial technology even if the size; of the national market as a whole may be large enough. Second, the cost of labor varies depending on how readily the labor force in rural area can move into the industrial sites. In one extreme where the supply of labor may tte infinitely elastic, as depicted in Arthur Lewis` hypothesis, the wage is maintained at the low level. But in the other extreme where the supply of labor is limited to the vicinity to industrial site, perhaps due to the high cost of migration, the wage may be too high to run the modern factory profitably. The conventional wisdom on the issue of the delay in industrialization has been pointing at the inadequate access to the advanced technologies, the lack of human capital, and the insufficient national resources to finance the investment in the industrial facilities. But the persuasiveness of such explanation is eroding away when we start thinking about the environments of the world economy today. Nowadays, the industrialization of an undeveloped economy does not require the invention of its own technologies because there is the fairly-well developed market for the modern, if not the frontier, manufacturing technologies. Since the transfer of technology is usually accompanied by the provision of managers and engineers as well as the standardized blue prints and manuals, the lack of human capital became a less insurmountable hurdle than it was in the early twentieth century. The development of the global network of financial markets in the late twentieth century made it much easier for an undeveloped economy to acquire financial resources necessary for industrialization. Furthermore, all of the above problems can be resolved, if an undeveloped economy succeeds in inviting foreign direct investments. Section 1 presents the basic model of an undeveloped economy. Section 2 introduces