This paper provides a glimpse into the nature of moderated and non-moderated French-language chat in public, non-educational contexts, and enumerates and explains a number of pedagogical applications aimed at helping learners of French participate in and have greater access to this type of authentic discourse beyond the classroom. It first presents a corpus-based analysis of language variation in moderated and nonmoderated chat discussions, focusing in particular on orthography, the second-person pronouns tu and vous, the first-person plural pronouns nous and on, and verbal negation, in order to demonstrate the amount and types of variation that exist in these two different types of synchronous computer-mediated communication. It then provides a number of pedagogical recommendations for teaching sociolinguistic variation in French using computer-mediated communication data. The sample tasks and task components are conceptualized to include one or more spheres of learning opportunities proposed by the New London Group.