This study starts with a simple question: how does an audience to a speech know when to applaud? In seeking to answer this question it draws on the ethnomethodological approaches of Atkinson (e.g., 1984a, 1984b & 1985), and Heritage and Greatbatch (1986) in order to study five speeches, each given to a live co-present audience by Barack Obama. On the basis of close analysis of transcribed data from the speeches, it seeks to identify the ways in which audiences successfully find occasions for collective applause. It concludes that many occasions for applause are sequentially related to certain types of patterning in the wording of the speeches and in this respect the current study confirms the findings of Atkinson (e.g., 1984a, 1984b & 1985), and of Heritage and Greatbatch (1986). However, this study also goes beyond their findings, in some important ways. The instances of applause elicited by verbal patterning (46.60%) occupy less than half of the total examples of applause in sharp contrast to Atkinson’s (1984a), and Heritage and Greatbatch’s (1986) findings. This suggests that other kinds of feature may also play a significant role in occasioning applause. This study also extends previous discussions in the literature in the following ways. Firstly, the function of projecting a name has been identified in more detail and has been further elaborated. Secondly, I argue that the irregular use of lists of three can mislead the audience as to when to applaud. Thirdly, many examples in this case study confirm that lists of two have particular effects on the production of applause. Fourthly, four other ways which do not depend on particular kinds of patterning but are commonly used in speeches, namely greetings, closing remarks, proposing applause for a named other and humor, are clearly related to applause elicitation in my data. Keywords: applause elicitation; political speeches; Barack Obama; rhetorical; non-rhetorical; ethnomethodology; conversation analysis (CA).