Soil/atmosphere exchange fluxes of nitrous oxide were monitored for a 3-year period at 10 grassland sites in eight European countries (Denmark, France, Hungary, Ireland, Italy, The Netherlands, Switzerland and United Kingdom), spanning a wide range of climatic, environmental and soil conditions. Most study sites investigated the influence of one or several management practices on N₂O exchange, such as nitrogen fertilization and grazing intensity. Fluxes were measured using non-steady state chambers at most sites, and alternative measurement techniques such as eddy covariance and fast-box using tunable diode laser spectroscopy were implemented at some sites. The overall uncertainty in annual flux estimates derived from chamber measurements may be as high as 50% due to the temporal and spatial variability in fluxes, which warrants the future use of continuous measurements, if possible at the field scale. Annual emission rates were higher from intensive than from extensive grasslands, by a factor 4 if grazed (1.77 versus 0.48 kg N₂O-N ha⁻¹ year⁻¹) and by a factor 3 if ungrazed (0.95 versus 0.32 kg N₂O-N ha⁻¹ year⁻¹). Annual emission factors for fertilized systems were highly variable, ranging from 0.01% to 3.56%, but the mean emission factor across all sites (0.75%) was substantially lower than the IPCC default value of 1.25%. Emission factors for individual fertilization events increased with soil temperature and were generally higher for water-filled pore space values in the range 60–90%, though precipitation onto dry soils was also shown to lead to high losses of N₂O-N from applied fertilizer. An empirical, multiple regression model to predict N₂O emission factors on the basis of soil temperature, moisture and rainfall is developed, explaining half of the variability in observed emission factors.