Wetlands are significant carbon (C) sinks and are expected to promote greater C assimilation as atmospheric CO₂ concentrations rise. However, the fate of C with environmental change along fresh-to-oligohaline wetland transitions is not well understood. We established an ex-situ mesocosm experiment to mimic future elevated atmospheric CO₂ concentrations (eCO₂, 720 ppm) versus current (380 ppm), and we exposed four co-occurring coastal wetland communities that naturally transgress (i.e., freshwater forest, mixed forest and marsh, marsh, mudflat) to these concentrations for two years. Overall, wetland communities with marsh plants in monoculture and mixed culture maintained high ecosystem C uptake with eCO₂ versus freshwater forested wetlands or mudflats, likely from superior plant species photosynthetic adjustment versus leaf area increases. eCO₂ promoted greater CO₂ uptake by leaves in all communities except mudflats, while promoting CH₄ efflux from whole ecosystems only when marsh plants were present. eCO₂ is projected to stimulate C gain 2.2-fold for forested wetlands and oligohaline marsh and 2.9-fold for forest-marsh mixture through greater CO₂ uptake. However, this comes at a cost of stimulated CH₄ flux by 1.4-to-1.7-fold in mixed and marsh communities versus reduced CH₄ fluxes with eCO₂ by forest and mudflat communities, perhaps through different oxidation pathways. Freshwater forested wetlands limited greenhouse gas emissions compared with transitional habitats, oligohaline marshes, and mudflats as atmospheric CO₂ concentrations increased. Stimulated C uptake in marshes may not offset higher methane emissions from these systems, potentially facilitating greater warming in a future with elevated atmospheric CO₂.