In no-tillage systems, crop residues are left on the soil surface and form a mulch, which can result in the stimulation of N₂O emissions through various processes. As a consequence, the accurate description of mulch decomposition and the associated N₂O emissions by soil-crop models is essential to help manage the potential impacts of mulches on the environmental performances of these agroecosystems. Here, we combined the use of a soil-crop model (STICS) and published experimental data to develop a new model representation of mulches (with various masses and qualities) and their decomposition, the carbon (C) and nitrogen (N) dynamics, and effects on N₂O emissions. We used a published dataset from southern Brazil combining two residues with distinct chemical characteristics (vetch, Vicia sativa L., and wheat Triticum aestivum L.) and four rates of mulch addition (0, 3, 6 and 9 Mg dry matter (DM) ha⁻¹) decomposing over one year. The STICS model with its default parameterization overestimated the remaining mulch masses, particularly at high DM inputs, and underestimated the N₂O emissions. The evolution of the STICS soil and decomposition modules led to two major results: i) we modified the mulch module parameterization, by representing that the whole mulch is accessible to decomposers, whatever its initial mass and thickness; ii) a new potential denitrification function was proposed, which uses simulated CO₂ fluxes associated from both soil humus and residue decomposition as proxy for C readily available to denitrifiers. With these new representation and parameterization, the model then accurately reproduced the very large range of magnitudes and the temporal variability of C and N fluxes observed for the two residues and the four mulch masses. These results are promising and the conceptual formalisms generic enough to be potentially developed in other soil-crop models. The next step is now to extend and generalize them to other conditions.