Despite arguments justifying the need to consider how cultural ecosystem services are coproduced by humans and nature, there are currently few approaches for explaining the relationships between humans and ecosystems through embodied scientific realism. This realism recognises that human–environment connections are not solely produced in the mind, but through relations between mind, body, culture and environment through time. Using affordance theory as our guide, we compare and contrast embodied approaches to common understandings of the co-production of cultural ecosystem services across three assumptions: (1) perspective on cognition; (2) the position of socio-cultural processes and (3) typologies used to understand and value human–environment relationships. To support a deeper understanding of co-production, we encourage a shift towards embodied ecosystems for assessing the dynamic relations between mind, body, culture and environment. We discuss some of the advantages and limitations of this approach and conclude with directions for future research.