Multiply substituted isotopic species of nitrous oxide (N2O), referred to as clumped isotopes, represent a promising new tool for distinguishing production pathways of this potent greenhouse gas. This work presents the first determination of enrichment factors of N2O clumped isotopes during bacterial denitrification. Samples of N2O obtained after 1-, 3-, and 7-day incubations of a pure culture of the denitrifier Pseudomonas aureofaciens at 20 °C and 30 °C were analysed by the recently developed quantum cascade laser absorption spectroscopy (QCLAS) method. Enrichment factors εp/s of the cumulative product (p) relative to the substrate (s) were determined using a Rayleigh model for the seven most abundant isotopically substituted molecules (isotopocules) of N2O. Values of the enrichment factors εp/s (with uncertainty expressed as expanded standard uncertainty at the 95% confidence interval) at the two incubation temperatures (20 °C/30 °C) are:14N15N16O (456): ε456 = (−40.3 ± 2.6)‰/(−35.1 ± 0.7)‰15N14N16O (546): ε546 = (−38.1 ± 3.4)‰/(−31.2 ± 0.6)‰14N14N17O (447): ε447 = (21.3 ± 1.2)‰/(24.5 ± 0.5)‰14N14N18O (448): ε448 = (38.8 ± 1.5)‰/(46.4 ± 1.2)‰14N15N18O (458): ε458 = (−8.9 ± 2.0)‰/(−11.7 ± 0.6)‰15N14N18O (548): ε548 = (−3.4 ± 1.1)‰/(−1.8 ± 0.5)‰15N15N16O (556): ε556 = (−85.9 ± 1.5)‰/(−63.9 ± 1.4)‰Temporal evolutions of the abundances of singly substituted N2O isotopocules during nitrate reduction agree with previously published experiments: there is normal isotope effect associated with the production of 14N15N16O and 15N14N16O; i.e., intermediates leading to 14N14N16O react faster than intermediates leading to 14N15N16O and 15N14N16O. However, the production of 14N14N17O and 14N14N18O is associated with inverse isotope effect; i.e., intermediates leading to 14N14N16O react slower than intermediates leading to 14N14N17O and 14N14N18O due to preferential cleavage of 16O during nitrate reduction to N2O. Isotopic fractionation at the incubation temperature of 30 °C was significantly lower compared to 20 °C. We observed a large kinetic isotope effect of the 15N site preference (SP) and the 15N–18O site preference (SP18) at the onset of the reaction. SP18 was found to be closer to 0‰ than SP, which is thought to arise from similar rates of breakage of the 15N–O and 14N–O bonds in the reaction intermediates. The 15N–18O clumped isotope anomalies in two isotopic isomers (isotopomers) 14N15N18O and 15N14N18O (Δ458+548avg) follow a temporal trend similar to those of SP and SP18. The 15N–15N clumped isotope anomalies in 15N15N16O are greater than 0‰ and show no clear temporal trend or influence of incubation temperature, suggesting no strong combinatorial effects involved during the N–N bond formation. Overall, our data illustrate that clumped N2O isotopes may be used as independent tracers for reaction mechanisms of N2O conversion and may establish themselves as a worthwhile tool to study the biogeochemical cycle of N2O.