Georeferenced information is growing every day, and geographical information systems are becoming crucial in many decision processes. As a consequence, extracting knowledge from GIS's may have an important impact. The paper presents a general approach for extracting sets of spatial transactions from GIS's, and for applying data mining algorithms to them. As a basic example of the process we present the extraction of spatial association rules from georeferenced data. Categories and Subject Descriptors: H.2.8 [Database Management]: Database Applications ? Data Mining, Spatial databases and GIS General Terms: Algorithms.