Site preference (SP), the difference in δ15N between the central and outer nitrogen atoms in N2O, is a powerful approach for apportioning fluxes of N2O from soils to nitrification and denitrification (Sutka et al., 2006). A critical aspect of the use of SP data to apportion sources of N2O to nitrification and denitrification is the need to evaluate data for isotope shifts that may have occurred during N2O reduction in soils prior to its escape to the atmosphere. We present data on the isotopologue effects during reduction of N2O during anaerobic incubation of soils and pure cultures of denitrifying bacteria. Isotopic enrichment factors for N2O reduction in soil mesocosms experiments varied between −9.2 and −1.8‰ for nitrogen and between −25.1 and −5.1‰ for oxygen. In pure cultures of Psuedomonas stutzeri and Psuedomonas denitrificans we observed isotopic enrichment factors for SP of −5.0 and −6.8‰, respectively. We further find that N2O consumption produces consistent relationships between δ18O and δ15N and δ18O and the δ15N of the central N atom in N2O of 2.5 and 1.6, respectively, which are clearly diagnostic of this process. Our results indicate that SP may be altered during reduction of N2O and thus bias evaluations of its origins. To understand the impacts of N2O reduction in soil flux studies on source isotope signals we modeled the isotope effects of N2O production occurring simultaneous with reduction and find increasingly curvilinear relationships between δ18O and δ15N and δ18O and δ15Nα with increased reduction. Consequently, a deviation from the linear mixing relationship between soil-derived and atmospheric N2O is an indication of extensive reduction. On the basis of our characterization of isotopic fractionation during N2O reduction, we show that the rate of reduction would have to be substantially greater than 10% of that of production to impact SP estimates of N2O from denitrification by more than a few percent. Nonetheless, reduction results in a small, but potentially important, increase in SP away from values proposed for bacterial denitrification (0‰) toward those associated with production from nitrification (33‰) (Sutka et al., 2006). On this basis, estimates of the proportion of N2O derived from denitrification obtained from SP values are underestimates and therefore conservative.