Placenames in Antarctica are uniquely complex. Antarctica’s distinctive political environment results in a situation where many national bodies confer names, but until recently there has been no coordination of this activity. The consequence of this is that many features are multiply named. The poor state of topographic mapping of the continent means that the majority of placenames are only defined by point locations, and in many cases the point location is based on old or inadequate mapping. In 1994 an Italian team started work on the Composite Gazetteer of Antarctica (CGA) on behalf of the Scientific Committee on Antarctic Research (SCAR). The first version was presented to SCAR in 1998 and is now the premier source of placename information for Antarctica. This work still continues, and will continue for the foreseeable future as errors are corrected and new content – in particular new descriptions – are provided by national bodies. The present database includes all names currently in official use in Antarctica. Unfortunately, the only geometry provided with the CGA is a point location for each feature, and this is unlikely to change for many reasons. This paper describes how an explicit hierarchical structuring of placenames can provide a geographic structure, and how this structure can assist name selection and placement.