A study was carried out to evaluate the effect of horizontal walking upon CO2 production rate by the carbon dioxide dilution rate technique. This was used as an indicator of animal energy expenditure. Two groups of three 18-month-old Aberdeen-Angus steers were assigned to two experiments. Average weights were 290 (s.e. 7·6) kg and 285 (s.e. 1·0) kg for experiments 1 and 2 respectively. Animals were allocated to individual pens and given 5·0 and 4·5 kg dry matter of a mixed diet for experiments 1 and 2 respectively. After a 45-day training period they were assigned to three walking treatments: 0 (T0), 3 (T3) and 6 (T2) km at 3 km/hfor 3 days in a Latin square design (3 × 3). 14C labelled sodium bicarbonate (5·4 μCi/h), diluted in carbonate-bicarbonate buffer sterile solution 0·1 mol/l, was infused for 92 h intraperitoneally with portable peristaltic pumps carried by the animals. The CO2 production rate was calculated as the ratio between the rate of infusion (μCi/h) and the specific activity of CO2 (μCi/ml CO2) in saliva samples, which were taken, in experiment 1, as an average of the day (09.00 to 16.00 h) and the night (16.00 to 09.00 h of the following day). In experiment 2 the day was divided as follows: prior to activity (09.00 to 13.00 h), activity (14.00 and 15.00 h) and post activity (16.00 h). CO2 production rate (ml CO2 per h per kg M0·75) at resting was 817 (412 kj/kg M0·75), increasing during walking to 1·46 of the resting level (T1 and T2, experiment 2) with no differences between the 1st and 2nd h of activity. One hour post activity, the CO2 production rate returned in T2 to the level of T0 but in T2 remained at 1·28 times that of T0. The average CO2 production rate during a complete day or night (experiment 1) was not affected significantly by the activity. Assuming that CO2 production rate during walking is 1·46 of resting (experiment 2) and remains at that level even at lower speeds, it can be estimated that a daily 6 km walk would increase resting energy expenditure from 1·04 when walking takes 2 h, as in this experiment (3 km/h), to 1·11 when the animal spends 6h(1 km/h).