Two experiments were conducted, each over several months, when cows grazed either ryegrass (September-November 2001) or kikuyu (February-March 2002) pastures, to assess the effects of accurately allocating feed on a daily basis to lactating Holstein-Friesian dairy cows. In each case, 28 cows were randomly stratified into 2 equal groups on the basis of milk and milk component yield, liveweight, age and days in lactation. The metabolisable energy requirements of the animals were estimated from standard established requirements. In each experiment, both groups of cows received the same amount of supplement over a period that was equivalent to a pasture regrowth cycle of 12-16 days. The control group received a set amount of supplements each day, while supplements fed to the adjusted group varied, dependent on pasture available. Available pasture was varied from 7 to 21 kg DM/cow.day (above a stubble height of 5 cm), to mimic the variation found on well-managed dairy farms. When pasture available was above the predicted requirement for cows in the adjusted group, pasture availability was restricted to predicted requirements and the extra milk that could be produced from the spared pasture was estimated. However, cows in the control group had the opportunity to eat more pasture if allocated more than required. This could result in more milk being produced, a gain in liveweight, and/or a higher post-grazing pasture residue (and hence potentially improve pasture regrowth). If less pasture than required was allocated to the control group, production could reduce or the cows might graze harder. Thus, in the control group the proportion of forage to supplement remained relatively constant, but intake varied in relation to pasture allocated, while for the adjusted group the total intake was kept relatively constant. In experiment 1 (ryegrass), the milk yield, percentage of milk fat and liveweight change of cows in the control and adjusted groups was not significantly different. However, the cows in the adjusted group produced 0.016 kg/cow.day more milk protein. As the control group ate 0.35 kg DM/cow.day more ryegrass pasture (P = 0.008) it is assumed that accurate daily allocation of feed improved feed efficiency. In experiment 2, the milk yield and percentage of milk protein of cows grazing kikuyu pastures was not significantly different between groups but the percentage of milk fat and covariate-corrected liveweight at the end of the experiment was higher in the control group than in the adjusted group. The pasture spared by cows in the adjusted group was predicted to produce 8.9% more milk when grazing ryegrass pasture and 12.3% when grazing kikuyu pasture. Linear regression analysis of pasture on offer on post-grazing pasture residue was not significant for the cows in the adjusted group but was significant for the control group cows when grazing either pasture, indicating success in accurately allocating supplementary feed to maintain a constant grazing pressure. The results of this study should assist dairy farmers in deciding whether the effort required to allocate feed accurately to dairy cows on a daily basis, is worthwhile.