English teachers are frequently asked to help colleagues prepare presentations for international conferences. Sometimes this assistance takes the form of a language course or tutorials focusing on the spoken language. Contact time is short but the participants are highly motivated; therefore, which features of the spoken language should the teacher focus on? What type of pronunciation work will provide the greatest payoff in terms of successfully being understood when speaking English to an international audience? Given the current debate on norms, varieties and intelligibility - spurred on by the work of Jenkins (2000, 2002, 2007) and other proponents of English as a Lingua Franca - how can teachers ground their course design in research? This paper addresses a variety of issues concerning the design of pronunciation courses which focus on maximum intelligibility for both native and non-native speakers. Reference is made to an exploratory study of a particular course for researchers in applied linguistics, in order to illustrate some of the issues. Directions for further research are described.