Nitrous oxide (N2O) fluxes were measured from small-scale farmland features which are not explicitly accounted for when compiling greenhouse gas inventories for livestock farming systems. These included poached land surrounding feeding and water troughs, seepage from liquid and solid manures, gateways, tracks and ditches. Measurements were made on two dairy and two beef-and-sheep farms with contrasting soil types located in the southwest of England, UK. High temporal and spatial variability of N2O fluxes were observed. Some features were persistent hotspots of N2O emission, and fluxes from all features were predominantly higher than those measured from adjacent pasture. Seepage areas from liquid manures and poached land around water troughs produced the highest N2O fluxes with maximum emissions of 288 kg N2O ha−1 yr−1 and 212 kg N2O ha−1 yr−1, respectively. The main soil property influencing N2O fluxes was nitrate concentration. The spatial extents of the features relative to the whole farms were small and therefore, despite the high fluxes measured from them, their contribution to the whole farm flux generally was not significant. However, on one farm the emissions from the farmland features were 14.3% of the IPCC estimated total farm flux, suggesting that current methodology could, on occasion, be underestimating N2O emissions from livestock farms.