The measurement of pasture output raises many problems, because harvesting is carried out by grazing stock, and yields cannot be measured directly, as with grains, roots, hay and silage. Pasture output is also a function of management, and seasonal yield is frequently of more importance than total annual yield.Methods used in pasture studies have been reviewed and recommendations made by a Committee on pasture research (17). These include the use of 'unit grazing days' [VINALL (35) ], total digestible nutrient output [KNOTT et al. (22)] and mowing techniques, as well as some of the techniques described below. While mowing methods frequently give good agreement with output measured as T.D.N.'s or Starch Equivalent, this is usually only found over prolonged grazing trials, when cumulative output figures are used. Thus LINEHAN et a/. (24) have shown that over a period of three years a clipping method has given an output 101 per cent of that obtained from animal data. During individual grazing periods, however, output from clippings varied from 66 to 252 per cent of animal output. This illustrates the difficulty of evaluation of output over short periods. Some of the reasons for this have been discussed by GREEN (13). who indicated the large numbers of herbage samples that are required to obtain an accurate measure of consumption, and stressed the importance of sampling herbage at a level lower than that to which stock will graze. However, under certain conditions and with the taking of large numbers of samples, useful output data can be obtained: thus WAITE et al.(36) have measured the output of'strip' grazing, under which the errors introduced by differential growth, discussed by LINEHAN (24) are largely eliminated. SEARS (31) has described a method in which herbage in 'protected' areas is plucked until it is similar in appearance to grazed areas, but here allowance for herbage growth during the grazing period is difficult, and the method is basically subjective. The use of a 'browse unit' method has been suggested for range conditions by COOK et al. (3).The use of animal output as a measure of pasture production requires considerable numbers of animals because of individual variations in maintenance and production requirements. Moreover it is uncertain how far standards for maintenance and production, obtained in indoor experiments, are strictly valid in the field. Its application to short grazing periods is doubtful: thus TAYLER (34) has indicated the difficulty in measurement of liveweight increase due to variations in 'fiir. Little is known also about the energy expenditure in grazing.In general, the precision of measurement of pasture output by mowing techniques or animal output data is inadequate and interest is now largely concerned with alternative methods of evaluation. Of these, methods involving collection and chemical analysis of faeces show promise, and are being widely studied. There appear to be two quite distinct approaches in use, namely the use of indigestible tracer substances in the feed and t...