It is often argued that today's large metropolitan areas have exceeded the optimal size and, therefore, development efforts should be directed toward smaller urban centers.1 This argument is frequently based on the proposition that the economic return from investment, either private, public, or both, in large metropolitan areas is less than that in mediumand small-size urban centers. The purposes of this paper are (1) to demonstrate that this proposition cannot be supported by any available empirical analysis and that, in terms of economic efficiency, even the largest metropolitan area in the world is likely to be less than the optimal size; and (2) to point out that there is a legitimate objective for pursuing a decentralization policy for urbanization, that is, interregional equity, whereas in practice a decentralization policy is frequently proposed and undertaken for economic efficiency, an objective which this policy seems to affect adversely rather than favorably. Many writers have observed that the currently large metropolitan areas already have excessive population concentration. However, their assertions are rarely consistent with an economic efficiency criterion in the sense of increasing a country's total income. If they are made on the economic ground, they are not based on a systematic quantitative analysis but on either visual impressions or partial quantitative data. Only recently, sufficient quantitative data have become available to make a fairly strong