We describe an experiment where cattle urine was applied at a rate of 420 kg N ha−1, equivalent to 10 L m−2, to mesocosm swards of a high lipid genetically modified perennial ryegrass line (HME) and its wild type (WT). We measured N pools and fluxes in the plant and soil, soil microbial populations and N2O production. HME plants produced 21% greater biomass than WT (p = .02), resulting in greater N uptake (27% higher in HME, p = .05). Urine N recovery in total plant biomass during the experiment in HME and WT swards were 54.7% and 33.9% respectively. The nitrification potential of soil was significantly lower (p = .01) in HME than WT. Partial least square‐discriminant analysis using microbial gene abundance data indicated that HME and WT plant growth induced distinct microbial populations in the soil. These differences in plant soil microbial interactions between HME and WT swards resulted in significantly lower N2O emissions from the HME sward. Total N2O emissions over the 4 weeks after urine application was 38% lower (p < .03) in HME swards than in WT swards. The next step in this work is to identify the specific changes in HME traits that drive the reduction in N2O.