Five ruminallyfistulatedAngus steers (360 (s.e. 15·4) kg) were givenfood to maintain body weight constant: maize silage (TO) at 0·6 kg dry matter (DM) per 100 kg body weight twice a day. After a 15-day adaptation period they were infused continuously (I) for 96 h, with a solution of NaH14CO3 at a rate of 7 to 8 micro Curies (μCi) per h. On the last 2 days of infusion 30 g (Tl) and 60 g (T2) urea, respectively, were placed in the rumen at the end of the morning meal. Spot samples of urine (250 ml) were taken before and 5 h after the morning meal and after at least 24 h of infusion. Thereafter, animals continued with TO for one additional week, in which they were prepared with catheters inserted in salivary ducts and infused for 48 h, as previously described. Eighteen pairs of spot samples of urine and saliva were takenfrom three of thefive steers (369 (s.e. 20·7) kg), over a period of 5 h, after at least 24 h infusion (six per animal). Rate of carbon dioxide (CO2) production was estimated as the ratio USA (specific activity of CO2)from which energy expenditure was calculated (22 kj/l CO2). Silage composition, in situ degradability and ruminal ammonia and pH were measured. In situ degradability in thefirst 6 h was 200 g/kg and ruminal ammonia was in the range of 20·6 to 39·6 mg/l. Ammonia increased rapidly to 394·2 (T1) and 673·9 mg/l (T2) 1 h after addition of urea into the rumen but in 6 h in situ degradability was unchanged. Ruminal ammonia decreased linearly at rates (mgll per h) of89·3 in Tl (R2 = 0·57, s.d. = 21·5) and 151·6 in T2 (R2 = 0·81, s.d. = 23·3). Animal energy expenditure rates were not affected (P > 0·05) by treatment (TO = 15·6, Tl = 15·6 and T2 = 15·8 kj/h per M075). There was no difference (P > 0·05) in CO2 production rate (mllh per kg M0·75) determined from the SA of CO2 from urine (604) or saliva (630) samples. It was concluded that the energy cost associated with detoxification of the excess of ruminal ammonia was of minor importance in terms of total animal energy expenditure and that estimations ofC02 ratesfromsamples ofurine or saliva are comparable.