Age composition of neighborhoods in urban area are well known to be variated. The purpose of this paper is to analyse both the spatial pattern of such variations and their changes from 1960 to 1975 within Sendai city.The procedure is as follows: firstly, neighborhoods are conveniently defined as small administrative districts (chos) within Sendai city. Twenty six neighborhoods are chosen for analysis as samples representing all of neighborhoods within Sendai city. Secondly, 26 neighborhoods are classified into a small number of clusters on the similarity of age composition, using the method of Cluster Analysis. Spatial arrangements of such clusters are mapped. Thirdly, spatial variations of age composition is compared with those of occupation composition in order to clarify the coincidences between them.The results may be summarized as follows.1) Twenty six neighborhoods were classified into four clusters on the similarity of age composition in 1960. At the same time, the spatial arrangement of four clusters formed a series of circles within Sendai city. That is, the first cluster is characterized by both comparatively high ratio of 15-24 age group and low ratio of 0-14 age group. This comprized only one neighborhood (Higashi-Nibancho) located in CBD. The second one characterized by relatively high ratio of 35-44 age group consisted of neighborhoods located in the zone around CBD (8 blocks). The third which had the relatively high ratio of 25-44 age group included neighborhoods distributed in the inner zone (the old built-up area excluding the above two zone, 11 blocks). The fourth was characterized by both comparatively high ratio of 0-14 age group and low ratio of 15-34 age group, contrasting with the first cluster. Neighborhoods comprized in this cluster spreaded themselves in the outer zone (suburbs, 7 blocks).2) In 1975 period, Sendai city was divided two regions which consisted of the extended CBD and the outer region. The differentiation between CBD and the zone around CBD distinguished in 1960 became to be diminished in 1975. The two zones were combined as the extended CBD. In addition, the differences between the south part of inner zone and the outer zone in 1960 also became a little, and the two zone were able to be grouped into the outer region.3) The same spatial pattern of age composition by neighborhoods were identified in the spatial variations of occupation composition by neighborhoods. So, the Multiple Regression Analysis had been done in order to clear up the relationship between these two spatial variations. As a result, spatial variations of the proportion of persons engaged in both trade and farming were clarified to be closely related to those of age composition in 1960. But in 1975, such a strong relationship between them was not clearly found. The result guessed that it was necessary to analyse on the scale of Sendai Metropolitan Areas. Because residential developments had been done beyond the administrative area of Sendai city after 1960 and areal differentiation had also developed on that scale.