The nitrogen-use efficiency (NUE) of a fertiliser has implications for pasture growth and the environment. This study aimed to compare application of urea as a foliar spray or in granular form, to kikuyu (Cenchrus clandestinus (Hochst. ex Chiov.) Morrone) and short-rotation ryegrass (Italian ryegrass, Lolium multiflorum Lam.) pastures in the subtropical dairy region of eastern Australia. The first experiment was a replicated grazing study on a site with a high plant-available soil N (75 mg nitrate-N/kg). The granular rate of urea was 46 kg N/ha.month equivalent, and the foliar spray rate was 40% of the granular rate. Pasture growth rate (51 DM/ha.day with foliar spray vs 45 kg DM/ha.day with granules) and pasture consumed (4942 vs 4382 kg DM/ha) were not significantly different between treatments. However, over the 8 months of the study, soil nitrate-N levels fell from 75 to 22 mg/kg on the foliar plots but only fell to 60 mg/kg on the granular plots. The second experiment was a replicated plot-cut experiment on a site with a low plant-available soil N (8.7 mg nitrate-N/kg). The NUE for kikuyu grass was similar for all treatments with a mean of 14.8 kg DM/kg N for the four foliar treatments (high and low, with and without wetting agent) and 17.4 kg DM/kg N for the granular treatment. The NUE for the ryegrass was also similar for all treatments, with a mean of 13.2 kg DM/kg N for the foliar treatments and 15.8 kg DM/ha for the granular treatment. A third experiment, evaluating absorption of foliar-sprayed urea over time, found that >80% of the urea applied to kikuyu was absorbed by 7 h; for ryegrass, the amount absorbed was only ∼45% but increased to ∼75% if wetting agent was included. We suggest that the lack of benefit in NUE achieved by applying urea as a foliar spray, which contrasts with results from studies in temperate dairy farm systems, is primarily associated with the substantially lower tiller density and hence the smaller canopy area for absorption of the foliar spray by the new regrowth shoots post-grazing.