Computer-Mediated Communication (CMC) is acknowledged to represent a social space where people interact with others who may not necessarily know them. They can also recreate their own identities in the course of their interaction. This study investigates ways in which the identities of speakers can be revealed by their use of language in multiple-participant conversations. In particular, the study aims to elicit the strategies that speakers employ the most by analysing the way they talk at a micro-analytic level, and the ways in which they organise and sequence their turns at talking. The results show that the processes of turn-taking and topic development are subject to distraction and breakdown in computer-mediated environments. There are many instances of pauses caused by frequent overlaps between participants. The accents of participants are considered the main feature which can constitute one’s identity in voice-based chat-rooms. Other factors such as communication and technical skills, systems and server speeds could also have an effect on such communication. Additionally, the participants seem to employ certain strategies to overcome interactional limitations of CMC systems, such as the use of pauses, quiet and loud intonation, and stress of particular syllables of some words. These strategies can contribute to determining the speaker’s identity.