1. A series of calorimetric experiments was conducted with sheep which had fleeces ranging in thickness from 0·1 cm. to 12 cm. at environmental temperatures between 8 and 32° C. Heat production, heat loss by radiation, by convection and conduction, by vaporisation of water and due to warming food and water to body temperature were measured together with losses of energy in faeces, in urine and as methane.2. The effects of a rise in environmental temperature on digestion of the food and on the loss of energy in urine or as methane resulted in a slight rise in the metabolisable energy of the ration by 6 Cal./° C.3. Environmental temperature had a marked effect on heat production, particularly when the fleece was short. The critical temperature (i.e. the environmental temperature at which heat production was minimal) of the closely-clipped sheep varied from 24° C. at a high level of feeding to 38°C. at a sub-maintenance level of feeding. These critical temperatures are similar to that of naked, resting man but much higher than that of the pig when fed similarly.4. As the fleece grew the critical temperature fell. Thus, on a maintenance level of feeding, a sheep with a fleece of 0·1 cm. had a critical temperature of 32° C.; when the fleece had grown to 2·5 cm. the critical temperature was 13° C. while with a 12 cm. fleece the critical temperature was 0° C.5. Below the critical temperature heat losses increase more rapidly in sheep with light fleeces. Thus a heavy fleece not only depresses the critical temperature but also reduces the rate of increase of heat loss with falling temperature under sub-critical conditions.6. At environmental temperatures well below the critical, the heat losses of the sheep per unit surface were identical. Under such conditions, when the whole of the metabolisable energy of the food is used to keep the animal warm, the criterion of ration adequacy is a high content of meta-bolisable energy in small bulk.7. At environmental temperatures above 32° C. the heat production on a constant ration increased, the rise being greatest with the highest level of feeding. Consequently the net energy value of the food declined at these high environmental temperatures.8. The calorimetric experiments were supplemented by two comparative feeding trials in which the effects of normal outdoor environmental conditions on the body weight of groups of Cheviot and Blackface sheep were measured. Control groups were kept indoors in heated pens.9. During the mild winter of 1956-7 the out-wintered Blackface wethers i n full fleece did not loose any more weight than those fed the same rations indoors.10. During the more severe winter of 1957-8, Cheviot, in-lamb ewes kept on a maintenance diet gained 2·3 lb.; those kept outside on the same ration lost 3·3 lb. With Blackface, in·lamb ewes the difference between the two groups was 0·3 lb. in favour of the indoor group.11. The food utilisation of sheep is affected considerably by environmental conditions. With little fleece the critical temperature is high and even when in full fleece an effect of cold can be demonstrated under practical conditions.