This chapter examines referential practices in a corpus of colloquial Indonesian conversation and attempts to address the question: “Does referentiality matter to speakers?” I take referentiality to be a multi-dimensional phenomenon involving (at least) whether referents are construed as general or particular and whether they are tracked through discourse. Through close examination of excerpts from conversational interaction, I show that there is often a blurring between the general and the particular, that referents are often indeterminate, and that referentiality as a discrete and classifiable linguistic property often does not seem to be relevant to participants in ongoing interaction. In this sense, referentiality does not always appear to matter to speakers. At the same time, referential practices do appear to be exploitable by speakers as resources for social action. Specifically, I show that a shift in referential practices regularly coincides with a shift in alignment, understood in terms of footing. These may involve shifts between generalising and exemplifying or moving from narratorial to quoted speech. To the extent that shifts in referentiality coincide with shifts in footing, then it can be said that referentiality does matter to speakers.