Reducing emissions from deforestation and degradation (REDD) should focus on places where such emissions occur. Protected Areas (PAs) are, in theory, protected and hence, should have no emissions associated with land use/land cover change. In practice protection is incomplete. Can PAs be included in REDD schemes? Can ‘paper parks’ be included that exist on paper rather than in reality? How concrete should threats be before we call carbon (C) protection ‘additional’? The dilemma may be more manageable if protected areas are included in a broader landscape approach to REDD. Some REDD project proponents currently focus on ‘buffer zones’ where protection is incomplete, but biodiversity co-benefits of additional C protection can be large. The results of a REDD feasibility appraisal in an area surrounding the Lamandau River Wildlife Reserve in Central Kalimantan, Indonesia’s REDD pilot province illustrate the challenges of finding synergies between sustaining livelihoods for local communities, protecting orangutans and globally appropriate mitigation actions.