In the context of planning, managing and generating territorial knowledge, Geoprocessing plays a fundamental role in supporting strategic decision making through the treatment of georeferenced information, which is the object of interest in public and private institutions and organizations. The structure of databases provided by the Geographic Information Systems (GIS) is configured as a cartographic information generator to support the various productive chains of the agricultural sector, leading the geographic intelligence along the processes of decision making of territorial nature. The State of Minas Gerais, as Brazil's main milk producer, demands dynamic production and sharing of geographic data and information. For this, matrix or vector physical thematic maps produced and arranged in a Geographic Database provide the basis for spatial analysis that, by exporting to GeoWeb environment, enables the consolidation of a Territorial Information System focused on the planning, knowledge and management of the milk production chain. This research work aimed to construct a Territorial Information System for the milk supply chain in the State of Minas Gerais, using a Geographic Information System (SIG), sites with downloadable geographic data and a WebGIS-designed tool hosted on the Milk Intelligence Center (CILeite) of Embrapa Dairy Cattle. The methodology consisted of gathering the various databases, mainly those originating from IBGE (Brazilian Institute of Geography and Statistics), along with retrieval, with necessary corrections, spatialization, with assignment of access key referring to the municipality, micro or mesoregion for the database and selection of layouts in the SIG, organization and definition of WebGIS for publication of geographic information via the CILeite for the State of Minas Gerais. For the publication of the results, the MapServer software was selected, which, although free of charge, demands expressive investment in its continuity, in view of the constant updating of the server and changes resulting from the project in which it was inserted. Thus, after gathering data on the municipal to mesoregional scale, a base was built that included production maps, milking cows and animal productivity, as well as Digital Elevation Model, vectors for providing data on infrastructure, hydrography and Conservation Units, all of which remained available on system web pages for download and viewing, browsing and editing in the SIG and Google systems such as Google Earth. The created geographic database was stored on the desktop at SIG and on the CILeite website, through which access areas were checked upon their launch, mainly due to the maps with selected layouts and the downloadable pages. Other free map server software can be used for map publishing and use of other available features, noting aspects of data traffic on the network.