What do speech reports tell us about the act being reported? When such a question is pursued in connection with reports of the form ‘S said that p,’ answers typically focus on the semantic content of the speech act. Indeed, there is a familiar line of research that aims to exploit our understanding of (the truth and falsity of) speech reports, in order to reach conclusions about the semantic content of sentences or expressions (see e.g. Evans, 1982; Kaplan 1989a, 1989b; Soames 1989; Heck 1995; though see Cappelen and Lepore 1997 for objections to this approach). In this chapter I want to focus attention on another matter: the illocutionary force of the act being reported. In particular, I want to argue that there is a use of speech reports of a related form (and involving the same verb ‘to say’), reflection on which can help us discern aspects of the force of the act being reported. The use I have in mind is what I call the buck-passing use of speech reports, as when one speaker, challenged to defend a claim or belief of hers, does so by reporting another speaker as having said so. The thesis of this paper is that the legitimacy of this practice depends on two key pragmatic features of the reported speech. This result can be seen as establishing a non-trivial desideratum for theories of the illocutionary force of the type(s) of act in question.