Some of the highest losses of nitrate to surface waters come from drained agricultural land. This article studies the effect of subsurface drainage density on nitrate losses and crop yield under Belgian farming conditions, using the nitrogen version of DRAINMOD (Brevé et al., 1997). DRAINMOD (Skaggs, 1997) was used to simulate the performance of the drainage system of the Hooibeekhoeve experiment, situated in the sandy region of the Kempen (Belgium), and this for a 14-year (1985–1998) period. In the analysis a continuous cropping with maize was assumed. Daily NO₃-N losses were predicted for a range of drain spacings and depths, two drainage strategies (conventional and controlled) and three fertiliser application dressings (225, 275, 325 kg N ha⁻¹). Losses of N in subsurface drainage occur almost entirely in the NO₃-N form. Losses of organic and mineral N in the form of NO₃-N in surface runoff are negligible. Hydrologic results indicated that increasing drain spacing or decreasing drain depth reduces drainage discharge while it increases runoff. The use of controlled drainage reduces subsurface drainage and increases runoff. Results also revealed that increasing the drain spacing or decreasing the drain depth reduces nitrate-nitrogen (NO₃-N) drainage losses and net mineralisation, while increasing denitrification and runoff losses. Controlled drainage caused a predicted reduction in drainage losses and an increase in denitrification and runoff losses. The optimal combination of drain density and management is one that maximises profits and minimises environmental impacts. Simulated results indicated that NO₃-N losses to the environment could be substantially reduced by reducing the drainage density below the level required for maximum profits based on grain sales. The study concluded that if the environmental objective is of equal or greater importance than profits, drainage systems can be designed and managed to reduce NO₃-N losses while still providing an acceptable profit.