Assessing the N2O fluxes balance is a key challenge to estimate the effect of agriculture practices on greenhouse gas production. N2O fluxes remained difficult to measure on a field scale due to high spatial and temporal variability and usually low concentrations. Our work aimed at (i) characterizing by laboratory measurements soil potential N2O emissions from nitrification and denitrification and (ii) testing a modelling approach of N2O emissions that circumvents the problem of discrete measurements for two Brazilian rainfed rice cropping systems, no-tillage (NT) vs. disk tillage (DT). This latter approach consisted in the combination of 2 models: a mechanistic water transfer model and a N2O emission model, namely PASTIS and NOE. Simulations with the PASTIS +NOEapproach showed for bothNTandDTtreatments that: (i) the soil emitted low amounts of N2O, (ii) emissions by denitrification corresponded to short periods of high N2O emissions (15 times as high as emission by nitrification), (iii) nitrification contributed to ca 35% of the total N2O emissions at the crop cycle scale, (iv) field N2O emission measurements corresponded to the low bound of simulated emissions from nitrification.