Adoption of land-based effluent treatment systems can be constrained by the costs and availability of land. Sufficient land area is needed to ensure nitrate leaching from applied effluent is minimised. One approach to decrease required land area is to enhance N removal by denitrification. Layers of organic matter (100 mm thick) were installed below topsoil of a site irrigated with dairy factory effluent. These "denitrification" layers were tested to determine whether they could decrease nitrate leaching by increasing denitrification. Four plots (10x10 m2 each) were constructed with a denitrification layer installed at 300 mm below the surface, and N losses were measured in leachate using suction cups every 3 weeks for 19 months. N in leachate was compared with 4 control plots. Denitrifying enzyme activity, nitrate concentrations, and carbon availability were measured in samples collected from the denitrification layers. These measurements demonstrated that denitrification occurred in the layer; however, denitrification rates were not sufficiently high to significantly decrease nitrate leaching. Total N leaching was 296 kg N ha(-1) from control plots and 238 kg N ha(-1) from plots with denitrification layers; a total of 798 kg N ha(-1) was applied in effluent. More than 50% of the leached N to 40 cm was as organic N, presumably due to bypass flow. Other studies have demonstrated that thicker denitrification layers (more than 300 mm) can reduce nitrate leaching from small-scale septic tank drainage fields but this study suggests that it is probably not practical to use denitrification layers at larger scales.