Urine patches in grazed pastures are a major source of the greenhouse gas nitrous oxide (N₂O). It is well-documented that the relative concentration of the various nitrogenous urine constituents varies significantly with diet. The effect of these variations on N₂O emissions from urine patches, however, has never been documented. In three incubation studies we (i) quantified the differences in N₂O emissions between urines of different nitrogenous composition, (ii) identified pathways leading to different emissions, and (iii) determined implications for future N₂O emissions studies. All three incubation studies used topsoil from a sandy pasture in the Netherlands, and measured N₂O and CO₂ emissions over 65 days. Occasionally, destructive analyses of actual denitrification, soil mineral N content and pH were performed. In the first study, we found a decrease in N₂O emission from 8.4% to 4.4% of urine-N when the hippuric acid (HA) content of (artificial) urine was increased from 3% to 9% of urine-N (consistent with a difference in diet). The second study confirmed our hypothesis that this was due to an inhibitory effect of the HA breakdown product benzoic acid (BA) on denitrification. Both N₂O emission and actual denitrification were reduced by approximately 50% when HA/BA levels increased. The third study quantified how accurately artificial urine of different composition reproduced N₂O emissions from real cattle urine of known-composition. Only the (often-used) replacement of HA with Glycine resulted in significantly higher emissions compared to real cattle urine. We conclude that (i) HA in urine acts as a natural inhibitor of N₂O emission through its breakdown product BA; (ii) this may be the basis of a novel diet-based mitigation strategy; and (iii) that artificial urine without HA is not a good replacement for real cattle urine.