1. Preface Traffic accidents in Japan are increasing recently in spite of the nationwide “traffic safety first campaign”. Unexpected accidents related largely to traffic accidents have been ranked at the fifth in Japanese death-order by causes, but in 1968 they jumped to the fourth order. The author has written a paper entitled “THE GEOGRAPHICAL STUDY OF TRAFFIC ACCIDENTS” in this journal, and now he would like to make a report focusing on Japanese metropolises. 2. Correlation of Traffic Stagnancy and Traffic Accidents in Metropolises For five Japanese metropolises (Tokyo, Osaka, Nagoya, Yokohama, and Kobe), correlation coefficients between the traffic stagnancy and traffic accidents were calculated by using the formula of correlation on the samples of Tab. 1. Results are as follows: (1) For the average of the five metropolises in the traffic stagnancy and the number of occurrence: r_??_0.64. (2) For the average of the five metropolises in the traffic stagnancy and the wounded: r_??_0.62. (3) For the average of the five metropolises in the traffic stagnancy and the dead: r_??_0.30. Next, taking Tokyo as an example among the five metropolises, the author made a correlation chart (Fig. 1) as to the relation of the traffic stagnancy and traffic accidents. In this figure, x is the traffic stagnancy, y1 is the number of occurrence, y2 is the number of the wounded and y3 is the number of the dead. And the following relations were found; y1_??_x+1.6, y2=_??_x+2.0, and y3=_??_x+4.2 As for such, if the traffic stagnancy comes to occur frequently because the traffic volume reaches over the limit of the metropolitan traffic capacity, both the number of occurrence and the wounded increase certainly in answer to it. While it comes to have the action of selfcontrol in traffic presently, thereby the traffic volume comes to decrease substantially within the metropolis despite of increasing number of registered motorcars externally. In addition, the motorcars which drive through the metropolis are forced to speed down because of the traffic stagnancy. These are why the number of the dead in the metropolis has been diminishing recently as shown in Tab. 2 and Fig. 2. Such circumstances within the metropolis are also why the relationship between the traffic stagnancy and the dead is flimsy on an average. 3. Regional Characteristics of Traffic Accidents within the Metropolis Tab. 3, Fig. 3, and Fig. 4 show the last decadal transition on traffic accidents in Tokyo Wards. Firstly, in the number of occurrence, the average of Tokyo Wards is 4.4 times. As compared with it, the center is 1.9 times, the middle part 3.3 times, the subcenters 4.8 times, and the fringe areas 7.3 times. That is to say, the center and the surrounding middle part are less than the average for Tokyo Wards in the increase rate. Above all, as Marunouchi of the center, there is even a region which showed a decrease of nearly 30%. On the contrary, the subcenters and the fringe areas are more than the average for Tokyo Wards. And there are not a few areas over 10 times in the decade as well, Above all, it shows about 22 times in Nishiarai area. For the dead, however, the average for Tokyo Wards shows a decrease of 6.1%. As compared with it, the center is a decrease of 51.3%, the middle part is a decrease of 17%, the subcenters show a decrease of 41.3 %, and only the fringe area shows an increase of 22.3%. Tab. 4, Fig. 5, and Fig. 6 show the distribution of the routes with the high density of traffic accidents per kilometer along roads. Consequently, it was seen that 27 routes of the high density in the number of occurrence distribute near the scope of the center and the subcenters generally.