(Kachru, 1992), including countries such as India, Pakistan and Sri Lanka, where a different variety of English is spoken (AMC, 2009(AMC, , 2011(AMC, , 2012a(AMC, , 2012b.In addition to the challenges of working in an unfamiliar medical system, these doctors can also face considerable difficulties in daily communications with patients, families, and colleagues (Hall, Keely, Dojeiji, Byszewski, & Marks, 2004;McDonnell & Usherwood, 2008;Tipton, 2005). In addition to the more obvious language difficulties with vocabulary, grammar, pronunciation, and an understanding of accents and colloquialisms, they can also struggle with the less salient but vitally important interpersonal features of language use, such as how to build rapport and show empathy (Hall et al., 2004;McDonnell & Usherwood, 2008;Pilotto, Duncan, & Anderson-Wurf, 2007). Moreover, many may not be familiar with the demands of the patientcentred models of care expected in developed medical environments in Canada and Australia, but less common in developing countries where medical facilities are very stretched (Dahm, 2011b;Khalil & Bhopal, 2009).In this article we focus on the communication challenges facing doctors who trained in medical environments very different from those found in Canada and Australia using a language other than English, in order to inform communications training designed specifically for doctors from language backgrounds other than English, and to illustrate how a close analysis of professional discourse can be transferred to ESL classes preparing for work environments beyond the medical world. We draw on clinical role-plays performed by practicing locally trained native English-speaking (NES) doctors and nonnative English-speaking (NNES) IMGs to identify the communication features of the kind of patient-centred approach to medical communication that will be expected of them. Although specific features and approaches to communication in Canada and Australia likely differ in some minor respects, our aims are to highlight features that are relevant in both cultures and to illustrate how discourse data can be used to identify culturally appropriate ways of communicating in a medical setting in order to provide an accurate evidence base from which culturally appropriate medical communication courses for IMGs may be developed.