Nitrogen (N) management is critical to the profitability of grain production systems, however careful management of fertiliser is needed to minimise environmental impacts. We investigated the effect of five N fertilisation strategies on nitrous oxide (N2O) emissions and nitrogen use efficiency (NUE) of rainfed wheat grown on a clay soil in a temperate, semi-arid environment of south eastern Australia during 2013 and 2014. Treatments included urea application (50 kg N/ha) at sowing with and without nitrification inhibitor (3,4–dimethylpyrazole phosphate) and surface broadcasting of urea with and without urease inhibitor (n-butyl thiophosphoric triamide) at the end of tillering plus an unfertilised control. Daily N2O emissions were low and responsive to in-season rainfall and fertiliser addition at sowing. Cumulative emissions from sowing until harvest were highest where N was applied at sowing in 2013; 160 g N2O-N/ha, while the 0 N control emitted 28 g N2O-N/ha (over 201 days). Emissions during 2014 were 77% lower than 2013 due to dry seasonal conditions; cumulative emissions were 49 g N2O-N/ha where N was applied at sowing, with background emissions of around 0 g N2O-N/ha (over 177 days). Inhibitors showed limited scope for reducing N2O emissions in this environment, however deferring N application until the end of tillering reduced N2O emissions. Grain yield responses to fertiliser were significant; increasing grain yield by 11–31% and NUE was generally high (recovery efficiency > 68%). However, deferring N application until the end of tillering in 2014 reduced yield (− 19%) and recovery of applied N (− 74%).