The absence of grazing management, coupled with limited N replacement, significantly promotes grassland degradation. Nitrogen input via fertiliser or biological fixation is critical for achieving productive and sustainable grassland systems. This two-year study assessed the N cycling of three pasture types (PTs): 1) mixed Marandu palisadegrass (Urochloa brizantha) and perennial horse gram (Macrotyloma axillare) without N fertiliser (Mixed); 2) monoculture Marandu palisadegrass fertilised with 138 kg N ha− 1 yr− 1 (N-fertilised); and 3) monoculture Marandu palisadegrass without N fertiliser (Unfertilised). Litter responses, forage and N intake, N livestock excretion and N cycling were measured. There were no differences between PTs for the existing litter (an average of 3,485 kg ha− 1; P = 0.204). More N was cycled via litter in the Mixed pasture than N-fertilised and Unfertilised pastures (95.7 vs. 63.6 and 51.8 kg N ha− 1 season− 1; P < 0.001). There was no difference between PTs for the total N excretion per animal (an average of 79.5 g animal d− 1; P = 0.727). However, per hectare, total N excretion in the N-fertilised pasture increased 20.4 and 22.5% than Mixed and Unfertilised pastures, respectively (P = 0.010). In the N-fertilised and Mixed pastures, there was a positive overall change of N in the soil-plant-animal system of 83.5 and 28 kg N ha− 1 year− 1, respectively. In the Unfertilised pasture, there was an overall negative change of N in the soil-plant-animal system of -27 kg N ha− 1 year− 1. Nitrogen application or perennial horse gram integration in a grass pasture increased the conservation of soil N reserves.