Aseries of trials was established in 2000, and measured until 2002, down the east coast of New Zealand on seven farms with contrasting amounts of spring-summer rainfall and resultant pasture growth. The aim was to evaluate the pattern of pasture responses to a range of nitrogen (N) fertiliser rates and associated soil phosphorus (P) levels generated by a range of P fertiliser treatments. Small plots excluded from grazing were established on steep (about 20 degrees slope) sites containing pastures that had been established for many years. Each of these steep sites was located within a hill paddock containing similar areas with a wide variety of slopes ranging from near flat to steep. Pasture production was measured approximately monthly (less frequently in seasons of slower growth), and soil was sampled for soil Olsen P test in mid spring each year. 90 kg N ha -1 ) were measured from three northern sites, whereas lower responses were recorded at the southern sites, and especially at Moeraki (3.7 kg DM kg -1 N applied). This may be partly explained by the lower temperatures in early spring at those sites. There was no pasture response over the late spring-autumn period in the North Island. However responses were measured at Moeraki in the South Island. There was little change in the average pasture response (kg DM kg -1 N fertiliser applied) to N fertiliser applied at 30 kg N ha -1 and at 90 kg N ha -1 at each site. This was in contrast to responses measured at the same locations on more gentle sloping sites where the average response declined as the rate of fertiliser N applied increased. Pasture responses in early spring to increase in soil Olsen P test were most pronounced on sites such as Wairoa and Wairarapa, with higher than average pasture clover content, and in years where the clover component was greatest. Some pasture responses to increase in soil Olsen P test also occurred over the late spring-autumn period, generally where there was significant clover content. Conclusions from these trials, together with those earlier reported, show that the initial grass:clover composition of the pasture, whether on easy or steep slopes, was able to be modified by both N fertiliser application, and by increasing soil Olsen P test, and was also a determining factor in the responses measured.