This paper explores the application of map generalisation techniques in the creation of small scale mapping (1:250,000) directly from large scale data (1:1250 and 1:2500). Small scale mapping provides a synoptic overview. At 1:250,000 scale, it is not that there is less information contained in the map, rather that we see fundamentally different information and relationships – broad-brush information between higher order phenomenon. The evaluation criteria used at this scale are quite different from those used at large scale. This paper describes a methodology by which small scale mapping is derived directly from the large scale (without generating intermediate results) using simple generalisation techniques. The approach illustrated in this case study, creates city boundaries based on simple buffering of building features, and derivation of the single polygon surrounding clusters of buildings over a threshold size. The hull is simplified, small internal polygons removed. The algorithm was applied to large and small conurbations alike. The results obtained via this method were evaluated against paper maps. Initial results point to refinement of this technique.