Using recently published data, a model of the sulphur (S) cycle for an established pasture, grazed by dairy cattle on Tokomaru silt loam, was constructed by assuming that organic S had reached an equilibrium level. The annual rate of net mineralisation, calculated in the model as 19 kgS hā¹, was the single most important contribution to the plant-available pool of S. In a field pot experiment involving growing plants under natural rainfall and temperature conditions, the amounts of S mineralised from 6 Tokomaru silt loam soils, differing in their organic matter contents, ranged from 10.9 kgS hā¹ to 22.9 kgS hā¹. The relatively high rates of S mineralisation in 2 of the soils collected from the most well-developed pastures suggested that these soils may be approaching the organic matter equilibrium assumed in the model. Such models may therefore be useful in predicting fertiliser inputs to these soils. In contrast, other soils under less well-developed pastures had very much lower mineralisation rates and are apparently still far from equilibrium. Highly significant relationships were obtained between the amounts of S mineralised in the field and the total S content of the soils. The accumulation of soil organic S within a soil type as indicated by the total S content may be a useful indicator of the approach to equilibrium and hence the extent of net S mineralisation. The presence of growing plants significantly enhanced S mineralisation in the field in most soils and reduced leaching losses of S from all soils. In conjunction with the field pot experiment, a long term field incubation and a series of shorter term incubations (both in the field and laboratory) were conducted to investigate the effects of temperature, moisture content, pH and soil pretreatments on S mineralisation. The amounts of S mineralised in the long term field incubation were found to vary markedly with time. In contrast, the actual levels of soil sulphate during the incubation period were less variable and were highly correlated with the amounts of S mineralised in the field pot experiment. If such incubation techniques are to be used, the amounts of sulphate at the end of the incubation may be a better indication of the ability of a soil to mineralise S in the field than the actual amounts of S released during the incubation. There were significant relationships between the final levels of sulphate at the end of all the incubation experiments and the total S content, again suggesting that within a soil type, total S content may be used to indicate the ability of the soil to mineralise S. In all incubation experiments there were negative relationships, for each soil, between the levels of sulphate initially present at the start of the incubation period and the amounts of S mineralised during that period. This indicates some type of 'end product regulation' which may involve sulphatase enzymes. No significant effects of temperature or moisture content were observed on the rate of mineralisation of soil organic S under the conditions of these incubations. The addition of lime was found to increase S mineralisation in all soils. The amounts of S mineralised after liming were significantly related to the pH attained in all soils, although in 2 soils they were better related to the amounts of lime added.