Dialectology addresses the study of the linguistic features of languages having a strong oral tradition like local dialects. These linguistic features can be of very different natures: phonetic, morphosyntactic, lexical, semantic or prosodic. They evolve according to space, time and/or social environment. To study local dialects, corpuses of phonetic data have been transcribed into Linguistic Atlases. These atlases gather a set of maps on which are registered, for a given lexical entry, the phonetic forms collected at several geographical points of inquiries. Between 1902 and 1910, J. Gilliéron and E. Edmont – precursors in the collect, description, the processing and the analysis of dialectal data – have built the French Linguistic Atlas (ALF). Made up of 1700 maps representing 639 points of inquiries, the ALF constitutes a corpus of 1.086.300 (1700x639) reliable lexical data. Such data were collected from a single questionnaire and noted in a homogeneous way with indication of places, dates and circumstances. This important document provides the base for the analysis of the lexical variations and the elaboration of second-generation atlases, called interpretative atlases, within the framework of international projects. Whereas the theoretical approaches used for the construction of linguistic atlas are structured, reliable and homogeneous, data analysis and the elaboration of interpretative maps is still based on manual and non-standardized approaches. One explanation for such a limitation is the fact that no software solution has been yet designed for the exploitation of geolinguistic data. Most of the work is indeed done manually which on the scale of the ALF is very time-consuming. In particular, the use of Geographical Information Systems (GIS) and spatial analysis methods is poorly developed with lexical maps that remain hand-drawn. Such limitations in the cartographic production capacities do not favour an efficient exploitation of geographical knowledge by researchers in dialectology. It becomes thus urgent to propose software that will facilitate the extraction, the analysis and the sharing of geolinguistic data. This paper introduces GIS for processing dialectology data. Since dialectological data are qualitative, unstructured and heterogeneous data, their integration into a usual GIS software raises various issues for: i) scanning and georeferencing old map; ii) geocoding point of enquiries; iii) structuring and modelling various components of geolinguisitc data; iv) creating geographical information layers adapted to the representation of geolinguistic data. We have defined the specifications of a GeoLinguistic Information System that provides researchers with spatial analysis and geovisualisation functionalities. The way how lexical and dialectological data - coming from heritage cartographical documents - can be integrated into a GIS are studied. We also study what kind of cartographical methodologies may be adapted to the treatment of dialectological data, and we propose various algorithms to automatically create “isoglosses”, namely limits that separate different linguistic areas. By taking into account the heterogeneous spatial distribution of geolinguistic point of enquiry, these algorithms allow to create homogeneous linguistics areas based on various dialectological criteria. The features of the dialectological data coming from Linguistic Atlases of France - edited more than 100 years ago - are presented. The methodology to structure and to model these data is described with their inclusion into a GIS. Our proposal is based on the design of geolinguistic information layers integrating phonetic, lexical, morphosyntactic, motivational and geographical dimensions. Finally, the main functionalities of our Geolinguistic Information System are described. We focus on the geovisualization of dialectological data and more specifically on the various ways to map isoglosses. A case study illustrates ours proposals.