There is a growing realization that the traditional approach to studying language, i.e. as a decontextualized,linguistic phenomenon, does not allow us to fully understand communication in the real world. Bystudying the way people process words and sentences in isolation, a wealth of variables that play a role innaturalistic communication are missed. In the study of aphasia, a language impairment caused by acquiredbrain damage such as stroke, a thorough understanding of the mechanisms of naturalistic communicationis imperative, as this is the behaviour therapy aims to improve. The field of aphasiology currently lacksa comprehensive, theoretically founded definition of communication. This lack of understanding, we willargue, makes it nearly impossible to accurately describe a person’s level of communicative ability in everydaylife as well as to predict with certainty what kind of intervention will lead to a change at the levelof communication. In this article we propose a model of situated language use borrowed from sociology,psychology, communication sciences and psycholinguistics, which covers both internal (e.g. individual) andexternal (e.g. environmental) factors that influence communication, including the traditional linguisticskills that have been extensively researched in the past. The model defines language use as: (1) interactive,(2) multimodal, and (3) based on context (common ground). An extensive review of existing researchon each component of the model in non-brain damaged adults and people with aphasia is provided. Theconsequences of adopting this approach to diagnosis and therapy for aphasia are discussed. The aim of thisarticle is to encourage a more systematic approach to the study of situated language use in aphasia.