Low phosphorus (P) constrains productivity of many of the permanent pastures of south-eastern Australia where wool and meat production are important. Whereas the need for fertilisation with P is indisputable for many soils, the benefits of liming to ameliorate soil acidification are not always clear in terms of forage production or quality. This experiment, by measuring botanical composition and groundcover over 10 years, aimed to address the issue of forage composition and groundcover. Merino wethers were continuously grazed on a permanent pasture under two stocking rates (low and high) following treatments with three rates of surface-applied lime and two rates of P. The lime rates were: nil, and sufficient to raise the pH_CaCl2 of the 0–10 cm soil layer to 5.0 and to 5.5. Phosphorus was added as two rates of single superphosphate: 125 kg ha^–1 every 2–3 years, and 250 kg ha^–1 every year. Two hypotheses were proposed: (i) as lime application rate increased, it would lead to the pasture swards having a higher percentage of species favourable for grazing animal production, and a higher level of sward groundcover would be maintained; and (ii) legume presence would be favoured by a higher level of applied P over the long term. The Millennium Drought (1997–2008), which coincided with this experiment (1998–2008), tended to mask or reduce lime effects; nevertheless, the higher legume content in the sward, as well as the superior groundcover found in the high lime–high P treatments at various times throughout the experiment and at its completion compared with the nil-lime control, indicated positive forage effects of lime application. Moreover, by the termination of the experiment in 2008, groundcover generally tended to be lower under nil lime than in treatments that had received lime. The higher proportion of legume in the sward was closely linked to greater applied P, although this did not occur at the low stocking rate in the absence of lime. By contrast, similar high P–low stocking rate treatments but with applied lime had the highest clover contents. It is proposed that the poor persistence of the treatment without lime may be due to a combination of poor legume seedset, because of excessive competition from accompanying grasses, and aluminium toxicity adversely affecting legume growth.