What effect does distance have on costs for economies at different locations? Exports and imports of final and intermediate goods bear transport costs that increase with distance. Production and trade depend on factor endowments and factor intensities as well as on distance and the transport intensities of different goods; The combination of distance, poor infrastructure, and being landlocked by neighbors with poor infrastructure can make transport costs many times higher for some developing countries than for most others. Drawing on two traditions of economic modeling - Heckscher-Ohlin trade theory and von Thunen's work on the isolated state - Venables and Limao analyze the trade and production patterns of countries located at varying distances from an economic center. Predicting a country's production and trade pattern requires knowledge of the country's location, its factor endowment, and the factor intensities and transport intensities of goods. Venables and Limao define transport intensity and show how location and transport intensity should be combined with factor abundance and factor intensity in determining trade flows. A theory based on only one set of those variables, such as factor abundance, will systematically make incorrect predictions. They report that geography and endowments interact in such a way that the world divides up into economic zones with different trade patterns. Countries close to the economic center may specialize in transport-intensive activities; countries further out become diversified, producing and sometimes trading more goods; countries still further out may become import-substituting (replacing some of their imports from the center with local production); in the extreme, regions become autarkic. More remote locations have lower real incomes. Globalization changes the terms of trade, improving the welfare of regions further out from economic centers, though reducing the welfare of closer regions. Where will a new activity, such as assembly of a new product, locate? Remote locations are disadvantaged if the product has high transport intensity (perhaps because of heavy requirements for intermediate inputs). But the costs of remoteness are already incorporated into the factor prices of those regions, which makes them more attractive. Which location is chosen depends, therefore, on how existing activities compare with the new activity in transport intensity and factor intensity. This paper - a product of Trade, Development Research Group - is part of a larger effort in the group to study the location of economic activity. The authors may be contacted at avenables@worldbank.org or ngl4@columbia.edu