Wildfires are expected to increase in frequency and severity across Canadian boreal forests under climate change, with important consequences for land–atmosphere carbon and nitrogen fluxes. Terrestrial biosphere models such as the Canadian Land Surface Scheme Including biogeochemical Cycles (CLASSIC) are used to quantify these impacts, yet wildfire emission factor (EF) parameterizations remain poorly constrained for Canadian boreal forests. Here, we assess the sensitivity of carbon and nitrogen pools in CLASSIC to EF parameterization and evaluate whether updating EFs using recent Canadian observations improves model representation of boreal wildfire carbon and nitrogen emissions. We conduct a sensitivity analysis using three simulations with prescribed fire at 15 Canadian eddy covariance flux tower sites. A baseline scenario using default EFs is compared to low and high EF scenarios, defined as one standard deviation below and above default values, respectively. Resulting changes in model outputs, particularly carbon and nitrogen pools, are evaluated and contrasted with model sensitivity to maximum carboxylation rate, a known influential parameter. In addition, we test a Canadian-specific EF parameterization derived from controlled laboratory burns by the Biomass Burning in Canada project at a subset of boreal eddy covariance flux tower sites. Simulated carbon and nitrogen emissions are compared to airborne observations of boreal wildfire plumes. Anticipated results include quantification of CLASSIC’s sensitivity to EF parameterization and assessment of whether Canadian-specific EFs improve agreement with observed wildfire emissions. These findings will inform interpretation of model outputs and support improved projections of wildfire impacts on climate and air quality in Canada.