Processing of geometric information is a major concern in applications such as computer-aided design of mechanical parts, computer-aided process planning, computer-aided manufacturing, and programming of industrial robots. Existing geometric models consist of rigid data structures of record-oriented databases with static relationships that often reflect a prime concern for efficient graphic rendering. This approach to geometric modeling restricts users to inadequate means of expressing and manipulating geometric concepts. A new approach is developed which will apply deductive reasoning for processing of geometric information at an abstract level within a geometric database. This database needs an object-oriented semantic data model tailored for geometric reasoning. The objective of this research is to develop a prototype geometric database with adequate reasoning capability and to study the fundamental concepts that must be incorporated in such a database to allow automatic inference of geometric properties. Logic programming and semantic data modeling techniques will be used for representation and manipulation of geometric information.