The author earlier made clear in his article (Kagawa, 1987) the areal distribution of the ageing phenomenon in each of the central cities in the six prefectures of Tohoku Region as well as its driving factors. He found in general that the census tracts with relatively high ratio of the elderly, or Aged Population Special Districts (APSD's), distributed in the central area of the city (core area and innner area) and in the urban periphery (Fig. 1 of this article). By using cohort analysis, it can be found that the main cause of ageing in the central area of the city was out-migration of the non-elderly, whereas in the urban periphery an absolute increase in the ageing population was as responsible to the phenomenon as out-migration of the non-elderly. Slightly different was the case of Sengai City, Tohoku's regional capital, where APSD's distributed only in the central area of the city.Is the case of Sendai applicable to other “regional capitals”? To find out the answer, the author selected Kanazawa City, Hokuriku's regional capital. To say conclusion first, Kanazawa's case was very similar to that of Sendai, where APSD's were found only in the central area of the city. This suggests that the metropolitan area of a regional capital can be pretty large in its spatial extent. In other words, the urban periphery of a regional capital is not rural area but urbanized area whose index of the increase in population is high. Then, in these areas, ratio of elderly population usually is relatively low, as the result of remarkable increase in non-elderly population. In the central area of the city, on the other hand, ratio of the elderly becomes higher, due to the salient out-migration of the non-elderly.In conseqnece, an areal distribution model of aged population, shown in Fig. 6, is obtained. It is explained as follows.Aaα: Due to heavy inheritance tax, small and/or old housing, etc., non-elderly population, tend to migrate out of this area, making the ratio of ramaining elderly population extremely high.Baα: Characteristics of population are very similar to Aaα.Baβ: While partly similar to Aaα, originally low ratio of elderly population still remains the same.Bcα: Increase of condominiums has brought new inhabitants into this area, while approximately the same number keep transferring out. Originally high ratio of elderly population ramains the same.Cbβ: Due to improper landform condition and poor commuting means, this area has not developed as residential area until recent years. But, it is developing as a new front, young nuclear families flowing in its area. Ratio of elderly population, therefore, is comparatively low.Ccβ: Owing to good landform condition and efficient commuting means, this area has been fully developed as residential area. Average age of inhabitants is slightly higher than those of Cbβ. This ares has good enough possibility that the whole area may turn into Caα or Ccα.Dbβ: Proper landform condition and efficient commuting means form residential area. As young nuclear families constitute a most portion of inhabitants, elderly population ratio is relatively low.Dcβ: Due to improper landform condition and insufficient commuting means, development of this area is fairly slow. Although, depending on how it is carried out, the area may become a part of Dbβ in future.Daα: Rural area. Continuous out-migration of the non-elderly makes the area one with high ratio of elderly population.Dcα: Rural area similar to Daα. Since the out-migration of the non-elderly has reached already its maximum, the area now faces the most serious ageing problem.