1. Three diets of barley–canula-seed (Brassica campestiris), lucerne (Medicago sativa) or chopped brome-grass (Bromus inermis) were given at intervals of 3 h to closely-shorn Suffolk wethers held at a temperature of 1–5° (cold) or 22–24° (warm). Apparent digestibility of organic matter (OM) and nitrogen was reduced by 0·08–0·05 and 0·04 units respectively for lucerne and brome-grass diets given to cold-exposed sheep, but no treatment effects on digestibility were observed for the barley–CSM diet. Measurements achieved using infusion of the digesta markers 58Co-EDTA and 103Ru-phenanthroline (103Ru-P) showed that cold exposure depressed apparent OM digestion in the stomach and intestines by 33 and 42 g/d for the lucerne diet, and 13 and 35 g/d for the brome-grass diet respectively.
2. The turnover time (h) of the l03Ru-P marker in the rumen of warm sheep was 38·9 for barley–CSM, 18·4 for lucerne, and 15·6 for brome-grass. In cold-exposed sheep, 103Ru-P turnover time (h) tended to be reduced to 32·3, 12·3 and 15·3 for the three diets, respectively. OM fermentation in the stomach was highly related to 103Ru-P turnover time for lucerne and brome-grass diets.
3. Cold exposure increased the escape of dietary N from the abomasum by 0·04 and 0·09 of dietary N intake for sheep given lucerne and brome-grass diets respectively. Dietary N degradation was closely related to 103Ru-P turnover time for lucerne, and to the proportion of large particles in rumen digesta for the brome-grass diet. Estimates of feed N degradation made by use of information on the rate of fermentation of the diet in nylon bags and 103Ru-P turnover time were consistently lower than those observed in vivo for barley–CSM and lucerne diets. Intestinal digestibility of non-ammonia N was not significantly changed by cold exposure.
4. Transfer of urea from plasma to the rumen was 1·4–2·5 g N/d for the barley–CSM and lucerne diets, but the value for brome-grass was 4·5–4·9 g N/d. Cold exposure did not affect urea transfer. The production of ammonia from feed and endogenous protein was approximately 0·66 and 0·47 g N/g N intake of barley–CSM and lucerne diets, with no effect of cold exposure. Cold exposure reduced the value from 0·57 to 0·38 for brome-grass.
5. The results are compared with those obtained previously with pelleted hay, and the importance of large particle breakdown in the prediction of OM and N fermentation using nylon bags is discussed.