Rising concentrations of nitrous oxide (N 2O) in the atmosphere are causing widespread concern because this trace gas plays a key role in the destruction of stratospheric ozone and it is a strong greenhouse gas. The successful mitigation of N 2O emissions requires a solid understanding of the relative importance of all N 2O sources and sinks. Stable isotope ratio measurements (δ 15N-N 2O and δ 18O-N 2O), including the intramolecular distribution of 15N (site preference), are one way to track different sources if they are isotopically distinct. ‘Top-down’ isotope mass-balance studies have had limited success balancing the global N 2O budget thus far because the isotopic signatures of soil, freshwater, and marine sources are poorly constrained and a comprehensive analysis of global N 2O stable isotope measurements has not been done. Here we used a robust analysis of all available in situ measurements to define key global N 2O sources. We showed that the marine source is isotopically distinct from soil and freshwater N 2O (the continental source). Further, the global average source (sum of all natural and anthropogenic sources) is largely controlled by soils and freshwaters. These findings substantiate past modelling studies that relied on several assumptions about the global N 2O cycle. Finally, a two-box-model and a Bayesian isotope mixing model revealed marine and continental N 2O sources have relative contributions of 24–26% and 74–76% to the total, respectively. Further, the Bayesian modeling exercise indicated the N 2O flux from freshwaters may be much larger than currently thought.