End of Project Report Lysimeter studies have shown the adverse effect of Fallow soil in releasing NO3-N to soil drainage water. A soil at Johnstown Castle under Fallow management gave mean NO3-N drainage water concentrations >MAC (maximum admissible concentration, 11.3 mg/l) in the 10 th , 11 th , 12 th and 13 th years of cultivation. In the 14 t h year this changed as the Fallow treatment showed a mean value MAC. On the same soil, barley receiving 120 kg N/ha fertiliser N, showed variable results - soil drainage water concentrations were MAC in another year; the mean values were the Guide Level. Winter wheat receiving 150 kg N/ha as fertiliser had all soil drainage water concentrations the Guide Level, depending on drainage water volume. These results apply to a soil in cultivation since 1985 having reduced organic N reserves. Higher NO3-N concentrations in soil drainage water would be expected with similar soils recently changed from grass to arable farming. On grassed lysimeters (Johnstown Castle soil), that had been growing barley for 10 years, a combination of 300 kg/ha fertiliser N with 126 kg/ha cattle slurry N, applied in December or February and the same amount of fertiliser N plus 120 kg/ha pig slurry N, applied in December or February gave soil drainage water samples that only breached MAC once in 12 samplings per treatment. The mean NO3-N concentrations were the Guide Level. Cultivating the soil, in order to re-sow grass, produced a large release of organic N via mineralisation and this combined with a small fertiliser N input (50 kg/ha) gave very high concentrations of NO3-N in the drainage water. A delay in sowing from June to the end of September exacerbated this problem. The final phase of experimentation showed very low levels of NO3- N in the drainage water which was primarily induced by decreasing fertiliser N input to 200 kg/ha and slurry N input to 50 kg/ha. Five soils, representative of major Irish soils, were subjected to lysimeter trials in a similar manner under grass. In the first experiment the soils received 300 kg/ha fertiliser N plus approximately 120 kg N/ha as pig or cattle slurry. In Years 1, 2, when mean values were pooled over treatments, MAC was breached 2, 3 times by Clonroche; 5, 3 times by Elton, 1, 2 times by Oakpark; 5 times and once by Rathangan soil drainage water. Applying cattle or pig slurry in December with fertiliser N, applied during the growing season, gave the highest number of water samples in breach of MAC. Reducing the fertiliser N to 200 kg/ha and the slurry N to 51 kg/ha drastically lowered the NO3-N concentrations in the drainage water to sustainable levels. Cultivation followed by Fallow for 3 months prior to sowing grass gave very high NO3-N concentrations in all soil drainage waters. Due to recycling of N via animal excreta, greater leaching of NO3- N is likely to occur on grazed grass receiving identical N inputs. European Union Structural Funding (EAGGF)