Vegetated buffer strips have been recognized as an important element in overall agro-ecosystem management to reduce the delivery of non-point source pollutants from agricultural land to inland water systems. A buffer strip experiment consisting of two tree species ('Eucalyptus camaldulensis' and 'Casuarina cunninghamiana') with two planting densities and a pasture treatment was conducted to determine the effectiveness of NO₃-N removal from a cattle feedlot effluent disposal area at Tullimba near Armidale, NSW Australia. Different management methods were applied for the buffers where grass and weeds were mowed 2-3 times during the second and third years and were not managed during the rest experimental years for the tree buffer, while grass was harvested 1-3 times per year for the pasture buffer. The differences between tree species and planting density significantly affected tree growth, but the growth difference did not significantly affect their capacities to reduce NO₃-N in soil surface runoff and groundwater. On average for all the tree and pasture treatments, the buffer strips reduced NO₃-N concentration by 8.5%, 14.7% and 14.4% for the surface runoff, shallow and deep groundwater respectively. The tree and pasture buffer strips were not significantly different in NO₃-N reduction for both shallow and deep groundwater while the pasture buffer strips reduced significantly more NO₃-N concentration in surface runoff than the tree buffer strips. Both buffer strips reduced more than 50% of surface runoff volume indicating that both the tree and pasture buffer strips were efficient at removing water and nutrients, mostly through a significant reduction in soil surface runoff volume.