Region is one basic unit of spatial analysis in GIS applications, which is a kind of entity existing widely in many fields such as environment science, land information science and so on. In general, the kind of entity has a common character of continuously spatial distribution, so there is inevitably fuzziness in spatial data used to represent its position when it is abstracted and generalized, and enters into GIS databases. Here, fuzzy set is taken in this paper as a strong and convenient mathematical bridge to model uncertainty and fuzziness under a GIS environment. Firstly, a new method of representing fuzzy region is proposed based on field model, and its spatial uncertainty is analyzed. On the basis of above discussion, we analyze some existing methods on how to describe the point set topology of fuzzy region by referring to some shortcomings and giving a kind of new description method. Thus fuzzy region's boundary, interior and exterior are defined in detail, and their geometric meanings are given. Finally, this paper ends with the discussion on how to describe uncertain topological relations caused by spatial uncertainty of fuzzy region based on the point set topology, and develops corresponding 9_intersection models. It is shown that the point set topology model proposed in the paper is an extension of the 9_intersection model by Egenhofer (1991).