Various urban problems by the overconcentration of the urban functions in Tokyo have been pointed out. As a countermeasure to long commuting time, the decentralization of some employment, has been proposed in order for commuters to commute from closer areas. Certainly, if jobs are matched with houses and the “shokuju-kinsetsu” urban structure is realized, the total commuting quantity will decrease. But, if commuters could not change their houses soon, the job-decentralization will cause the longer commuting time and quantity. In this paper, we examine how much the commuting time in Tokyo Metropolitan Area could be decreased at certain optimum. Applying Transportation Problem of Linear Programming to Tokyo Metropolitan Area, we compare and examine the minimized commuting time with the real one. In the existing studies, the standard form of Transportation Problem was applied to these problems of commuting time. But, this application has some points at issue. It is pointed out that the standard form of Transportation Problem, that has only a set of constraints that fix the total quantities from origins and those to destinations, is so simple and the solutions are so unrealistic. This can be related to the mathematical characteristics of the Transportation Problem. There are few “non-0-elements” in the solution if the number of the constraints is fewer than the number of the elements in the solution. In this paper, therefore, we solve the Transportation Problem by adding more plausible constraints based on industrial classifications, and compare the minimized commuting time with the actual one. Consequently, we find that the excess commuting has increased in all models between 1975 and 1995, and conclude that the job-decentralization could not effectively bring about the “shokuju-kinsetsu” urban structure.