In tropical dairy production systems, where high rates of urea fertiliser are applied, little is known about nitrogen (N) fertiliser response, fertiliser-use efficiency and losses to the environment. This study aimed to determine the effects of N fertiliser (urea) application rate and a nitrification inhibitor (3, 4-dimethylpyrazole phosphate, DMPP) on pasture yield and N losses in a dairy production system in North Queensland, Australia. The experiment was a factorial design with two fertiliser rates (industry standard per application 57 kg N ha–1 or half that, applied ~3-weekly), two DMPP rates (0 or 4.86 g kg–1 urea) and four replicates, completed over 1 year. Urea applied at half the industry standard rate, together with DMPP, provided annual dry matter pasture yields (11 462 kg ha–1) not significantly different from those when urea was applied at the higher rate, with or without DMPP (10 691 and 11 156 kg ha–1, respectively). The low rate of urea without DMPP had the lowest annual dry matter yield (8386 kg ha–1). Most of the fertiliser N lost from the system appeared to be via leaching, with loss in surface runoff minimal. During the ryegrass phase, an experiment with 15N labelling showed that, 3 months after application, 27–39% of the applied N had been taken up by the pasture, 23–45% was recovered in the soil, and 18–40% had been lost. Emission of N2O peaked within a day of fertiliser application, and DMPP did not reduce emissions during that period. The findings indicate good potential for farmers to use DMPP-treated urea as a means of reducing N fertiliser rates without loss of productivity and with less loss of N to the environment.