Isotopic signatures of N2O are increasingly used to constrain the total global flux and the relative contribution of nitrification and denitrification to N2O emissions. Interpretation of isotopic signatures of soil-emitted N2O can be complicated by the isotopic effects of gas diffusion. The aim of our study was to measure the isotopic fractionation factors of diffusion for the isotopologues of N2O and to estimate the potential effect of diffusive fractionation during N2O fluxes from soils using simple simulations. Diffusion experiments were conducted to monitor isotopic signatures of N2O in reservoirs that lost N2O by defined diffusive fluxes. Two different mathematical approaches were used to derive diffusive isotope fractionation factors for 18O (epsilon18O), average 15N (epsilonbulk) and 15N of the central (alpha(-)) and peripheral (beta(-)) position within the linear N2O molecule (epsilon15Nalpha, epsilon15Nbeta). The measured epsilon18O was -7.79 +/- 0.27 per thousand and thus higher than the theoretical value of -8.7 per thousand. Conversely, the measured epsilonbulk (-5.23 +/- 0.27 per thousand) was lower than the theoretical value (-4.4 per thousand). The measured site-specific 15N fractionation factors were not equal, giving a difference between epsilon15Nalpha and epsilon15Nbeta (epsilonSP) of 1.55 +/- 0.28 per thousand. Diffusive fluxes of the N2O isotopologues from the soil pore space to the atmosphere were simulated, showing that isotopic signatures of N2O source pools and emitted N2O can be substantially different during periods of non-steady state fluxes. Our results show that diffusive isotope fractionation should be taken into account when interpreting natural abundance isotopic signatures of N2O fluxes from soils.