Language, as an ostensive communication system, necessarily relies on an ability to recognise informative and communicative intent (Leslie & Happé, 1989). The ability to recognise communicative intent has been argued to depend on an ability to process fourth-order intentionality (Sperber, 2000), and pragmatic interpretation to depend on even higher levels of intentionality (Sperber, 1994). To date, research has shown that adults are not able to process higher than fourth-order intentionality (Kinderman et al., 1998; Stiller & Dunbar, 2007), which would render this account of ostension and pragmatics implausible, but which had important methodological flaws, and did not explore the promising avenue of implicit tasks found to be successful in child research (e.g. Baillargeon et al., 2010). This research tested the ability of adults to process up to seventh-order intentionality, compared it to the ability to process non-mentalising recursion, and tested the difference made by implicit vs. explicit presentation of tasks. Results showed that adults are able to process up to seven levels, although with increasing difficulty; that implicit presentation results in greater accuracy and ease of this task; and that non-mentalising recursion is processed with equal accuracy, but increased perceived difficulty and no effect of implicit presentation. These results suggest that adults are highly skilled at processing metarepresentation, especially in naturalistic contexts, lending plausibility to Sperber’s accounts of the origins of ostension and pragmatics (Sperber 1994, 2000).