Bottom-up wildfire inventories commonly assign fixed emission factors (EFs) to vegetation classes, introducing uncertainty because EFs vary with fuel conditions. Here, we conducted laboratory burns of live-origin and litter-derived dead <em>Pinus massoniana</em> needles across a broad range of dry-basis moisture contents (MC) to quantify stable-spread-stage EFs for CO₂, CO, and CH₄ and develop fuel-condition-resolved parameterizations. Burning was smoldering-dominated, with sustained flaming observed only for litter-derived dead needles at 0.89 % MC. Increasing MC decreased EF(CO₂) and increased EF(CO) for live-origin needles, whereas the opposite responses occurred for smoldering-dominated litter-derived dead needles. The moisture-response regressions for EF(CO₂) and EF(CO) explained 60.3&ndash;81.5 % of the variability in mean EFs, while EF(CH₄) showed no detectable moisture dependence and was represented by fuel-state-specific means of 15.7 &plusmn; 4.3 and 8.0 &plusmn; 2.1 g kg^-1 for live and dead fuels, respectively. Applying these parameterizations in a bottom-up wildfire emission framework showed that similar total gaseous carbon emissions can mask substantial shifts in species-level carbon partitioning. Over 2012&ndash;2021, cumulative gaseous carbon emissions differed by only 4.3 % from a fixed-EF baseline, whereas CO₂-C fell by 14.0 % and CO-C and CH₄-C rose by 154.6 % and 344.1 %, respectively. These shifts translated into an approximately 65.63 % greater 20-year global warming potential-based near-term climate impact. These findings highlight the importance of fuel-condition-resolved EFs for better representing wildfire emission composition and its atmospheric and climate implications.