Very high resolution (0.65m) satellite imagery, field enquiry about cropping systems and other georeferenced information are analysed in order to provide with maps of interest for the definition of geographical indicators to help delimiting a "Terroir" Arabica in the region of Kintamani in Bali. lt results in the production of severa! maps including: a) map of the main agrosystems practiced in this region: food crops, forests, systems associated with dense shading by large trees (Erythrina, Leucaena and Albizia), systems based on perenn ial crops (Ciove, Cocoa, Citrus, associations of Citrus and Arabica coffee, Arabica coffee cropped without shade), b) map of the main agroregions distinguishable in this area (agriculture dominated by food crops, by large trees shading, by Citrus and associations with Citrus, and by Clove), and c) Digital Elevation Mode! (DEM) of the region. These different products allow to implement spatial analysis of the Arabica coffee-based agrosystems distribution in the region of Kintamani in Bali and to derive q uantitative data about the areas planted with Arabica in this region. These products constitute the basis of a Geograp hical lnformation System ( G IS), enabling the superimposition of various ancillary data l ike sensorial quality characteristics, location of Arabica coffee processing places or topographie levels, to lead a spatial analysis of the relationships between these independent factors. This approach shows that some characteristics are c learly l inked with the production of good Arabica coffees. They are located inside the central agro-region dominated by Citrus associations (and more precisely at West of the line U lian 1 Dausa), and cropped only above the altitude limit of I 1OOm. The agrosystem used for Arabica coffee growing (e.g. shading by large trees or by Citrus) seems to have no effect on its qual ity. Nevertheless, it seems that no trivial and restrictive limits can be extracted for the "Terroir" of Kintamani Arab