With rapid urbanisation, the demand for office floor space also increases. Although high-rise buildings are very common in urban areas, study of the rent for different stories in such buildings is relatively rare. Liu (1988) and Grimaud (1989) separately applied agglomeration economies to study the density variation in urban areas. However, these studies assumed that the floor rent in the same building is the same for each story. Such an assumption about floor rent may not fully reflect actual floor rent and, from the perspectives of theoretical completeness and practical usefulness, formulation for different floor rents in high-rise buildings, and their effect on urban density variation should also be considered. This research note extends Ogawa and Fujita’s locational potential function (1980) into two dimensions: flat dimension and story dimension. The modification of the locational potential function is then applied to revise Liu’s model. Finally, the office distribution in a CBD (central business district) is then discussed.