1. One of the biggest difficulties in the measurement of the feed, intake of grazing animals by the indirect method using indigestible ‘markers” is to determine the optimum frequency for dosing and for taking faeces samples in order to achieve greatest accuracy while yet retaining realistic free-grazing conditions of husbandry.

2. Results of earlier studies of the rate of excretion of individual doses of marker by grazing sheep are used to illustrate the probable outcome of various schedules of dosing.

3. It appears that if marker doses were administered and faeces samples were taken twice daily at alternate intervals of 9 and 15 hr. approximately, the mean marker concentration of these samples would yield an essentially unbiased estimate of the general level of marker in the faeces. Since there is no satisfactory evidence that animals fail to excrete all the marker administered, it should be possible then to derive estimates of mean faeces output with an error probably of the order of + 10%.

4. This technique was tested in two stall-feeding trials and and one free-grazing trial with a total of seventeen sheep.

5. The unweighted mean of the estimates of faeces dry-matter output for the seventeen individual sheep periods was about 100 ± 8%.

6. Alternative methods of bulking the faeces samples so as to yield estimates of group mean faeces output are considered, but the possibility of introducing a serious bias by the bulking process is pointed out.