Coastal forested wetlands are globally important components of the carbon cycle. These ecosystems, and the carbon they store, are at risk of sea level rise and saltwater intrusion (SWISLR), which can kill tree stands and create ghost forests. Although ghost forests have clear consequences for aboveground carbon storage, how this ecosystem transition affects belowground carbon is less clear. Initial predictions on the effect of SWISLR on carbon emissions from ghost forests suggested methane emissions would decline due to the influx of sulfate in sea water. However, studies to date have found no consistent effect of SWISLR on soil carbon emissions in coastal forested wetlands. We designed a field survey of ghost forest sites to examine how forest mortality and saltwater intrusion impact carbon dioxide and methane emissions in different environmental conditions. We collected soil samples from 17 paired sites in coastal North Carolina, each consisting of one intact forest and one nearby ghost forest. We incubated the soils in anoxic conditions and measured carbon dioxide and methane production over seven days. Further, we explored key soil characteristics like conductivity and organic matter to deduce drivers of carbon emissions in these ecosystems. We found that both carbon dioxide and methane emissions were highly variable, with forest mortality having no consistent effect. Of the 34 individual sites, only six produced notable quantities of methane, distributed amongst live and dead forests, dry and inundated conditions, and across the salinity gradient. The best predictors of carbon dioxide emission rates were soil organic matter and reduced iron, while methane was weakly correlated with sulfide and elevation. Our results suggest that more research is needed to determine how carbon cycling in coastal forested wetlands is affected by forest mortality and SWISLR, and what this could mean for future climate forecasting.