1. Metabolism of urea in non-pregnant and pregnant sheep (1–25 d from term) has been examined. Injections of (14C) urea were used to estimate urea entry rate, urea pool size and urea space in sheep given 1000g of a diet of equal parts of crushed oats and chaffed lucerne hay (day a) and in the same sheep 4 d after the ration had been reduced to 250 g (day b). 2. On both experimental days (a and b), mean pool size was greater (14% on day a, 29% on day b) and urea space was greater (54% on day a, 24% on day b) in pregnant animals than in non-pregnant animals; mean plasma urea concentrations were lower (35%) in the pregnant animals on day a but were not significantly different on day b. 3. The entry rate of urea was similar in all the animals on day a, but was significantly higher (34%) in pregnant than in non-pregnant animals on day b. There was a significant decrease in urea entry rate in both pregnant (33%) and non-pregnant (86%) animals on day b. 4. The rate of excretion of urea was lower (26% on day a, 35% on day b) in pregnant animals, indicating a higher (31% on day a, 40% on day b) rate of degradation of urea in the digestive tract of pregnant as compared with non-pregnant sheep. 5. Measurements of urea entry rate have been used to calculate the upper limit of amino acid deamination in pregnant and non-pregnant sheep, and this has been used as an indication of the potential availability of amino acid carbon for glucose synthesis. It is suggested that, at a maximum, amino acids may have contributed the carbon required for 63 g/d and 52 g/d of glucose on days a and b respectively.