ABSTRACTFor three weeks after emergence, micro‐swards ofLolium rigidumwere defoliated to a height and at a frequency which represented a range of set stocking rates of less than, similar to, and more than 7–8 sheep/ha as based on the results of Greenwood and Arnold (6) and referred to as D1, D2and D3respectively. The effects of these defoliation regimes on the growth and development of emergent swards were evaluated according to their effect on increments of DM from plant parts, rate of appearance of leaves and tillers, concentration of soluble carbohydrates, and on the uptake and concentration of inorganic nutrients.Defoliation began five days after emergence, when the dry weight of the shoot was 3g/m2. After 20 days of treatment, 6±7 g/m2of DM had been removed from D1, 5±5 g/m2from D2and 51 g/m2from D3; and the dry weight of shoot remaining was 13, 7 and 4±5 g/m2on D1, D2and D3respectively. The effect of defoliation on dry weight of plant parts increased in severity in the order leaf 1, stem+sheaths, leaf 2, roots, leaf 3, tiller 1, tiller 2, leaf 4. Average relative growth rate of the sward was reduced from 17% per day on D1to 13% per day on D3. There was no necrosis of root tissue. Emergence of the later leaves and of the tillers was retarded by up to three days by more frequent defoliation. Alcohol‐soluble carbohydrates in the root fell from over 6% to about 1±5% on all treatments. Uptake of N. P and K was limited by increasing intensity of defoliation and because concentrations of those elements were not greatly affected, the limitation could be ascribed mainly to the effects of defoliation on plant size.The plausibility of the technique for the experimental simulation of grazing of emergent annual grass swards in Western Australia is discussed. It is concluded that for set‐stocking rates of less than about 12 sheep/ha, grazing, while reducing DM production just after emergence, is unlikely to produce severe physiological stresses i