We describe a submodel to simulate NOx and N2O emissions from soils and present comparisons of simulated NOx and N2O fluxes from the DAYCENT ecosystem model with observations from different soils. The N gas flux submodel assumes that nitrification and denitrification both contribute to N2O and NOx emissions but that NOx emissions are due mainly to nitrification. N2O emissions from nitrification are calculated as a function of modeled soil NH4+ concentration, water-filled pore space (WFPS), temperature, pH, and texture. N2O emissions from denitrification are a function of soil NO3- concentration, WFPS, heterotrophic respiration, and texture. NOx emissions are calculated by multiplying total N2O emissions by a NOx:N2O equation which is calculated as a function of soil parameters (bulk density, field capacity, and WFPS) that influence gas diffusivity. The NOx submodel also simulates NOx emission pulses initiated by rain events onto dry soils. The DAYCENT model was tested by comparing observed and simulated parameters in grassland soils across a range of soil textures and fertility levels. Simulated values of soil temperature, WFPS (during the non-winter months), and NOx gas flux agreed reasonably well with measured values (r2=0.79, 0.64, and 0.43, respectively). Winter season WFPS was poorly simulated (r2=0.27). Although the correlation between simulated and observed N2O flux was poor on a daily basis (r2=0.02), DAYCENT was able to reproduce soil textural and treatment differences and the observed seasonal patterns of gas flux emissions with r2 values of 0.26 and 0.27, for monthly and NOx flux rates, respectively. ¿ 2001 American Geophysical Union