There are two viewpoints in recognizing regions in geography, homogeneity and functional organization. It is necessary, in seeking spatial structure or regional structure, to investigate not only from an aspect of homogeneous regions but also from an aspect of functional or nodal regions.<br> Recently multivariate analyses(especially factor analysis and principal component analysis) have been employed in geography due to the development of computer and remarkable progress in statistics. As a consequence of these developments, objective and synthetic analyses have become more common in researches. Since the structure of homogeneous regions based on regional characteristics and the structure of nodal regions based on inter-regional flows are understood to a certain degree, attention is now directed to interdepen-dency between the two structures and the spatial structure as a whole. The purposes of this paper are firstly to reveal the structure of homogeneous regions based on socio-economic characteristics (Table 1) and that of nodal regions based on interregional journey-to-work, and secondly to examine the interdependency between the two structures. The study area is Nagoya Metropolitan Area (Fig. 1). The method of this study is similar to the frame-work of Berry's general field theory of spatial behavior. The main results are as follows:<br> 1) Structure of homogeneous regions based on socio-economic characteristics<br> By factor analysis of socio-economic characteristics of 59 unit areas, eight factors account-ing for 85.68% of the total variance were extracted (Table 2), and interpreted as follows: (1) commercial activities (34.08%), (2) manufacturing activities (16.31%), (3) white collar employment (9.13%), (4) agricultural activities (8.19%) (5) population dynamics (6.45%), (6) age structure of population (4.70%), (7) urban land-use (3.60%), (8) employment in secondary and tertiary industries (3.22%).<br> Factor one (commercial activities) has the highest percent of explanation, but the other seven factors also have specific meanings. In the spatial arrangement of fac toe scores, concentric spatial patterns for the 1st, 3rd, 5th, and 7th factors were readily recognized (Fig. 2-a_??_h). On the basis of factor scores on the eight factors, the 59 unit areaa were integrated by Ward's hierarchical classification. The integrated regions showed a concentric pattern (Table 3 and Fig. 3). That is, it became clear that the structure of homogeneous regions based on socio-economic characteristics in the Nagoya Metropolitan Area was mainly prescribed by the eight factors mentioned above, and spatialy there was a distinct concentric pattern centering around Naka-ku.<br> 2) Structure of nodal regions based on interregional journey-to-work<br> The interregional journey-to-work matrix (59, 59) was subjected to R-mode factor analysis, and eighteen factors were extracted (Table 4). On the basis of factor loadings (_??_|0.4|) and factor scores (_??_|1.0|), the 59 unit areas were grouped into 19 nodal regions by Goddard's method (Fig. 4-a_??_f). Each nodal region is composed of continuous unit areas and overlaps partially with other regions. Five nodal regions which were closely connected with the central city Nagoya were identified. Consequently, the structure of journey-to-work regions is characterized by a large number of small regions and the stratification of the regions.<br> 3) The interdependency between the structures of homogeneous regions and of nodal regions<br> To summarize the interdependency between the structures of homogeneous regions and of nodal regions, canonical analysis was applied to factor score matrices of socio-economic characteristics (59, 8) and interregional journey-to-work (59, 18). Three statistically significant canonical variables were synthesized. A canonical variable isolates an aspect of complex spatial covariation of many regional phenomena by decomposing and reintegrating them.