This dissertation proposes a new method of studying pragmatic language use in autism spectrum disorder (ASD). The existing research into pragmatic language use in ASD has predominantly used interviews and experiments in clinical environments to gather data from participants with ASD. These research methods focus heavily on comprehension, on comparing the performance of the participant with ASD to their non-autistic interlocutor. The methodology proposed in this thesis involves using observation and a naturalistic environment to gather conversational data that can be used to fill several gaps in the current research on pragmatic language use in ASD. In this study, people with ASD interacted with either non-autistic people or other people with ASD in triadic conversation sessions. These sessions were recorded, transcribed, and analysed for various pragmatic uses of language that previous research had observed were impaired, such as types of non-literal language and conversational turn taking. The pragmatic language use of three participants with ASD, representing three different levels of language ability, was analysed. The results showed that some participants used types of pragmatic language differently based on with whom they were communicating and also that some purported pragmatic impairments in ASD may not be as ‘impaired’ as they seemed based on previous research. For example, while all participants in this study used more interruptions when interacting with others with ASD compared to when interacting with non- autistic people, the rates of interruption were far less than what has been observed between non-autistic speakers (Hancock & Rubin, 2014), even though people with ASD have been described in the research literature as interrupting frequently (Ochs, Kremer-Sadlik, Sirota, & Solomon, 2004). The research presented in this dissertation has implications for future clinical research into pragmatic language impairments in any speech community and provides an additional methodology that can be used, expanding on the types of research questions that can be investigated in this area. / Thesis / Doctor of Philosophy (PhD) / This thesis proposes a new method of studying pragmatic language use in people with autism spectrum disorder (ASD). This methodology involves using observation and the data that can be gathered using it can fill several gaps in the current research on pragmatic language use in ASD. In this study, people with ASD interacted with either non-autistic people or other people with ASD in triadic conversation sessions. These sessions were recorded, transcribed, and analysed for various aspects of pragmatic uses of language. The pragmatic language use of three participants with ASD, representing three different levels of language ability, was analysed. The results showed that some participants used types of pragmatic language differently based on with whom they were communicating and also that some purported pragmatic impairments in ASD may not be as ‘impaired’ as they seemed based on previous research.