Nitrogen (N) fertiliser was applied to dairy pastures at nominal rates of 0, 200, or 400 kg N/ha per year on farmlets stocked at 3.24 (LS) or 4.53 (HS) Friesian cows/ha from June 1993 to June 1995. By June 1995, clover contents on farmlets LS200 and LS400 were 10.6 and 2.2%, respectively, compared with 16.8%onfarmletLS0. Most loss of clover on farmlets LS200 and LS400 occurred during the first 6 months of the trial when slow early-spring growth of clover combined with high N responses put clover at a competitive disadvantage. N had a smaller effect on clover content under the high stocking rate because of improved utilisation of N-boosted pasture. By June 1995, there was little difference in clover content between farmlets HS0 (15.4%) and HS200 (14.9) although clover content was lower on farmlet HS400 at 6.8%. The decrease in clover content under N fertiliser on the low-stocked farmlets was confirmed by the decrease in clover plant density and stolon biomass. N fixation activity was consistently higher under no N fertiliser than on either of the low-stocked farmlets which received N. Application of 400 kg N/ha per year had a significant (P < 0.05) effect on clover plant morphology but plants on farmlet LS200 were no different to those on farmlet LS0. Plants on farmlet LS400 had lower stolon length and stolon dry weight (DW), developed fewer axillary buds, and therefore had fewer stolons. Although total leaf DW and number of leaves/plant were lower on A95062 Received 27 September 1995; accepted 21 December 1995 farmlet LS400, values for number of leaves/stolon and individual leaf DW were not significantly different. There were no significant differences between populations in either root number/plant or root DW/plant.