[abstract:] Although the relative paucity of social cues in computer-mediated communication poses problems of the organization of social relations in cyberspace, recent studies have begun to focus on the ways in whicht this deficit is managed. This article contributes to this research by addressing the question of how participants distinguish between contexts in online discourse. Data on cues, and on naming practices in particular, in text-based virtual realities called MOOs illustrate the structure of contexts and the dynamics of contextualizing communication and interaction in cyberspace.