The Military Topographic Institute in Dobruska is the agency providing topographic data to the Army of the Czech Republic. Building of a GIS project called DMU25 started in 1994. Therefore the institute was equipped with four WS HP715 and software package ArcInfo release 6.1.1. The GIS DMU25 is a vector database with the same spatial precision as the military topographic maps 1:25000. Its extent will be more than 70,000 square kilometers - the whole territory of the Czech Republic. The DMU25 database is organized in tiles of size of one sheet of map 1:25000. Graphic information is stored in 20 coverages. As to the feature attributes, feature attribute tables contain only feature ID and feature type code. All remaining attributes are stored in the standard INFO database. Joining graphic information and attribute tables is made by relations. The database definition, i.e., a list of all topographic object types and their registered attribute types, is described in �The Catalogue of the Topographic Objects� derived from FACC (DIGEST). The structure of data combines the requirements on the collection of the current Czech topographic map's contents, requirements for maximal compatibility with the structure of FACC, and the requirements for generating of coverages in ArcInfo. This vector database includes not only information for the military GIS applications, but complete contents of the topographic maps as well. One of the applications is creating of topographic maps of scale 1:25000 and 1:50000. This solution is built on the automated map image generating from the DMU25 vector data. Entire software system is created in the AML macro language. It uses a user friendly menu for communication with the operator. Generating of the topographic maps is divided into four following logical steps: 1. Selection of data from DMU25 The goal of this block is to select information from DMU25 in order to display it in the topographic map 1:25000 and 1:50000. In case of the map 1:50000 it is necessary to join four 1:25000 workspaces first and than to solve the matching of joining object attributes. Searching for the inconsistency between adjacent maps (1:25000) is performed automatically, and then solved by the operator. 2. Map content reduction and object displacement The goal of this block is reducing of geometric map content with respect to rules of cartographic generalization. The solution is to select features that should be in the map and simplify their shapes depending on the map scale. The object displacement of linear features is necessary especially in the drainage layer and in communications layer. The reason is, that the map symbols are considerably larger than the feature in the real world. Therefore it is essential to move one of the features if they are too close to depict both of them. The collisions are searched for automatically and then solved by the operator, too. 3. Symbolization The goal of this block is to replace the definition sets of points by appropriate cartographic symbols. Plotting of the map including the frame marginalia and annotation is performed automatically. 4. Map annotation The annotation is generated interactively and the operator uses information from attribute tables. The operator interactively selects the attributes of a given feature from the database and then he places this annotation to the best location. The annotation is stored into an annotation coverage. All the generated cartographic symbols and annotations are placed into a final map composition. The result of this process is a postscript file that is used for generating of four films (separated color in CMYK) in a laser image setter. The final stage is an offset printing. In near future we prepare to insert all the topographic data into single ArcStorm database. This will not only allow to perform global GIS tasks, but it will also facilitate creation of maps that have map sheets and coordinate system different from original topographic maps.