Digestibility trials were conducted at two stages of lactation with each of 107 Friesian cows using three diets (60:40, 75:25, 90:10 proportions of compound feed to medium quality hay fed long). Each cow received throughout lactation fixed daily allowances of one of these diets at either 2-26 or 2-63 times average maintenance requirements (MM), or was fed ad libitum, average intake equalling 3-07 MM. The digestibilities of the diets were also measured on immature cattle at 0-76 and 1-33 MM and sheep at 0-65 and 1-27 MM.For all classes of stock the digestibility of dry matter increased linearly and that of fibre decreased curvilinearly with increasing proportion of compound in the diet. There was no interaction of effects of diet composition and amount consumed on digestibility of feed. The digestibility of both dietary components declined for the young cattle and sheep with increased intake. The digestibilities for lactating cows were lower than for other stock but did not change with amount consumed or milk yield. Digestibility was slightly higher for first and second calf cows than for adult cows; and was higher for organic matter in mid lactation compared with early lactation for adult cows.These results for dairy cows conflict with the generally accepted interpretation that a linear decline in digestibility occurs from once to five times maintenance intake, embracing all classes of stock. a s t n e proportion of compound increased in the diet the digestibility increased. The majority of experiments to study the effects There are several constraining factors regarding of amount consumed and diet composition on this interpretation that indicate the necessity for a apparent digestibility of feed (here referred to as re-opening of this question. First the above experidigestibility) by dairy cows have followed the mental method does not include a rigorous separation hitherto accepted procedure of variation in amount of the effects of the three main components: yield of of compound in the ration proportionally to yield: milk, amount eaten, and diet composition. Also the the higher-yielding cows received a greater propor-alliance of non-lactating to lactating stock is fretion of compound and probably had a greater dry-quently variable (see Discussion). These comments matter intake. Such trials in combination with on experimentation are strengthened by a group of digestibility trials on young cattle or sheep at much experiments, the results of which indicate little or no lower feed allowances expressed as multiples of decline in digestibility with increase in intake by maintenance (MM) than allowed to the lactating dairy cows per se. These trials largely, but not stock led to the conclusion that as the amount totally, involve rationing by systems other than consumed increased the digestibility declined, and 'feeding according to yield'.