Lumpiness of production factors within a country might overturn the predictions for the structure of trade by the factor-abundance (HO) model. Trade patterns, as predicted by this model, can both be magnified or reversed by uneven concentration of production factors within a country. Cities are the most characteristic manifestation of lumpiness of production factors and as a consequence different patterns of urbanization between countries might cause trade patterns to differ from HO predictions on the basis of the overall availability of production factors. We argue that urbanization indeed affects trade patterns. The consequence of this result is that urbanization should be included in empirical trade analysis; urbanization could, e.g. to the understanding of the missing trade puzzle.