A remote control sampling technique was developed for the collection of oesophageal fistula samples from goats. Number and size of samples can be varied and collected throughout the day without disturbing the animal's normal feeding behaviour. The equipment developed and tested in this study consists of an oesophageal fistula valve which allows the fistula to be opened and closed, a rechargeable battery pack and motor to operate the valve, a portable radio and receiver to activate the valve motor, and a harness to attach the equipment to the body of the animal In addition, a closing device to effectively seal large oesophageal fistulae (> 1 050 mm²), which in turn is required to accommodate the valve, was developed. During field tests with the fistula valve, 10 % of I 027 sampling attempts failed due to blockage of the valve, and an average of 1.3 incidences of equipment failure were recorded per animal per sampling day, from an average of 9.9 extrusa collections per day. Observed feeding behaviour (grazing/browsing) as well as grass / bush ratio of fistula valve and standard fistula bag samples of four goats, formed the basis for the evaluation of this technique. In addition, extrusa recovery rates, measured under controlled conditions, were used in the evaluation of this sampling method Differences in extrusa composition between the fistula valve and fistula bag techniques varied substantially during the browsing period within a camp and also between camps. Furthermore, during high frequencies of observed grazing, there were large differences between the fistula valve and fistula bag methods. During this study, the fistula valve technique provided a more realistic estimate (R²=.91) of the observed feeding behaviour of goats than the fistula bag technique (R²=.63). Under controlled conditions, the large oesophageal fistula, with or without the valve, enables high and consistent extrusa recovery rates (87 % recovery; SD 7.5). / Thesis (Ph.D.)-University of Natal, Pietermaritzburg, 1993.