Previous research on the use of head nods in talk-in-interaction has demonstrated that they can be used for various interactional purposes by speakers and recipients in different sequential positions. This report examines speakers' uses of nods in “third position“, in the course of “minimal post-expansions“ (Schegloff, 2007). It identifies three possible distinct types of nods. The first of these can be used to register a prior utterance as news; the second appears to be designed to register receipt of a prior utterance without treating it as news; and the third embodies features of the first two types, and may be designed to register receipt and acknowledgment of “dispreferred“ news. These findings are suggestive of rich complexities in the use of head movements in the production of actions-in-interaction, and of the importance of a fine-grained analytic approach for understanding their situated uses.