The effect of four commonly used dairy farm management systems (treatments), on nitrogen leaching to 1 m was studied over a 4-year period from October 2001 to April 2005. The treatments were (i) grazed plots receiving dirty water, (ii) 2-cut silage plots receiving slurry, (iii) grazed plots and (iv) 1-cut silage plots receiving slurry. All plots had fertiliser N applied; the soil was free-draining overlying fissured limestone. Mean 4-year N input (kg/ha) was 319 and mean annual stocking density was ~2.38 LU/ha. The annual average and weekly NO3-N and NH4-N concentrations in drainage water were analysed for all years, using a repeated measures analysis. For the annual NO3-N data, there was an interaction between treatment and year (P < 0.001). There were significant differences (P < 0.05) in NO3-N concentrations between the treatments in all years except the third. For the NH4-N data there was no interaction between treatment and year or main effect of treatment but there were differences between years (P < 0.01). Mean weekly concentrations were analysed separately for each year. For NO3-N, in all years but the third, there was an interaction between treatment and week (P < 0.001); this occurred with NH4-N, in all 4 years. Dirty water was significantly higher than grazed-fertiliser only and 1-cut silage in NO3-N concentrations in 2001–02; in 2002–03, dirty water and 2-cut silage were significantly higher than the other treatments; while in 2004–05, dirty water and grazed-fertiliser only were significantly higher than the other two treatments. The overall 4-year mean NO3-N and NH4-N concentrations were 8.2 and 0.297 mg/L, respectively.