Abstract The responses by hill pastures of widely varying land slope, aspect, and grass species composition to nitrogen (N) fertiliser application in May and August 1980 and August 1981 were examined. The yields varied widely for control pastures of all slopes and differing grass species composition. In contrast, pasture responses to N application were much more uniform, with the greatest responses occurring in spring, and being obtained on easy, ryegrass-dominant areas. Although appreciable residual effects of N fertiliser were measured after the initial responses in spring 1980, net depressions occurred in spring 1981. This effect was greatest on steep slopes, and may be related to the lower soil N status of the steep slope trial site. Simple correlation analysis was used to relate total pasture yield from N (N-treatment yield) with pasture and soil parameters measured at the start of each trial period. Significant correlations of N-treatment yield were measured with land slope, both ryegrass and brownt...