Advances in GPS and wireless networking technologies have enabled a new class of location-aware applications, including location tracking [10,2], location-enhanced messaging [3,9], location-based gaming(www.botfighters.com), and navigation aids for the visually impaired . However, these applications typically represent places quite simply, as a geographical point or a point plus radius. We conducted an experiment that showed that this simple representation is not expressive enough to represent the full range of people's everyday places. We also present a set of more complicated physical shapes that our subjects found sufficient to cover their places. These results identify representational requirements for location-aware systems, have implications for systems that aim to acquire place representations, suggest enhanced applications, and open up interesting avenues for future research.