Abstract
This study has the following purposes. The first is to clarify the influence of spatial characteristics of the central commercial district of Kyoto City on the illegally-parked bicycles, by using Geographically Weighted Poisson Regression. The second is to clarify the spatial variability of this influence, illustrating the results by using Geographic Information System. As a result of this study, we have found out that both point characteristics related to surveillance and territoriality and area characteristics related to accessibility and removal risk statistically have a significant influence on the illegally-parked bicycles. Furthermore, we have clarified that this influence on the illegally-parked bicycles fluctuates spatially, according to a local condition.