ABSTRACIExperiments are reported from tussock grassland soils in which different grasses were grown as spaced plants in existing turf, sown in cultivated and uncultivated soil, and grown in pot cultures.Differences are demonstrated among grasses in response to sulphur when nitrogen supply is improved. Cocksfoot and ryegrass showed a response to applied sulphur fertilisers when sown after cultivation, whereas resident grasses, principally sweet vernal and browntop, showed no gain in yield from sulphur after cultivation. Effects of cultivation and herbicide treatment on the availability of sulphur from different sources are also reported.Experiments are reported which indicate that the respo.,se to sulphate by cocksfoot depends on the level of nitrogen uptake. Where sulphate supply was very high relative to available nitrogen. there was little or no improvement in cocksfoot growth. At moderate levels of added nitrogen, treatment with high levels of calcium sulphate frequently resulted in significantly less growth than treatment with lower levels of calcium sulphate. Where nitrogen uptake is increased without improvement in sulphate supply, a large proportion of nitrogen in herbage may be in non-protein form.Results are discussed in relation to the likely trends in pasture composition and fertiliser needs in the course of improvement of grasslands in sulphur-deficient areas.