A convenient way of measuring the energy which grass or indeed any food supplies to an animal is as its metabolizable energy. Metabolizable energy is defined as the heat of combustion of the food less the heat of combustion of the faeces, the urine and the combustible gas, mostly methane, which are produced from it. It is usually expressed as kcal metabolizable energy/g dry food and the definition is strictly in terms of measurable quantities, measurement of methane involving use of a respiration chamber. The calorific value of the dry matter of grass, or the organic matter of grass, is subject to relatively little variation. Hutton (1961, 1962), for example, found the calorific value of pasture grass to range from 4-32 to 4-56 kcal/g dry matter, a variation largely associated with the protein content of the grass. The main determinant of the metabolizable energy of grass per unit dry weight or per unit organic matter is undoubtedly the loss of energy in the faeces. Faecal losses of energy rarely fall below 15% but can rise to 60% of the heat of combustion of the grass. Loss of energy as methane rarely exceeds 10% of the energy of the grass and loss of energy in urine rarely exceeds 9% of the energy of the grass. Losses of energy both in urine and as methane vary over a considerably smaller absolute range than does the faecal loss of energy. Indeed, the metabolizable energy of grass and grassland products can be estimated with little error by multiplying the apparently digested energy of grass (heat of combustion of grass less heat of combustion of faeces) by 0-82. This factor applies to artificially dried grasses varying in protein content from 5-1 to 22-6% (Armstrong, Blaxter & Waite, 1964), to a dried lucerne (Bateman & Blaxter, 1964), to grass silages (J. L. Clapperton, 1963, unpublished) and to fresh pasture grass (Blaxter, Ekern & Sawers, 1963, unpublished). Alternatively, but with a slightly greater error, the metabolizable energy of grass can be estimated by assuming that each g digested organic matter has a metabolizable energy value of 3-6 kcal.Even so, such indirect methods of estimating metabolizable energy still involve a biological test in the form of a digestion trial. In our laboratory attempts have been made to estimate the metabolizable energy of grassland products from their chemical composition alone, commencing with a study of a series of artificially dried pasture grasses (Armstrong,