Intensive dairy farming with low N use efficiencies may have adverse environmental impact through nitrate leaching. The residual effects of six different temporary grasslands (1994–96) on yield and nitrate leaching in the following cereal crops (1997–99) were investigated on a loamy sand in central Jutland. The grasslands were unfertilized grass–clover and fertilized ryegrass subject to cutting or continuous grazing by dairy cows with two levels of N in feed supplements. In the first year there was sufficient residual effect of the grazed grasslands to obviate the need for supplementary fertilizer, but in the following years gradually more fertilizer N was required to obtain optimal yields. Nitrate leaching decreased as a function of time after cultivation of grassland, but grassland management had little effect on the subsequent nitrate leaching (6 to 36 kg N/ha in unfertilized plots). Application of cattle slurry to cereals influenced nitrate leaching more than the history of the grassland and caused the annual mean nitrate concentration to exceed the EU Drinking Water Directive upper limit in most cases. Presumably, large differences in N-input during the grassland phase of the crop rotation had relatively little effect on the subsequent N release because of variable N losses during grazing. Possibilities for further improvement of the utilization of grassland N following cultivation are limited when the current knowledge has been implemented. If the N use efficiency of dairy farming systems is to be further improved the utilization of N during the pasture phase is crucial.