The effect of walking on cattle energy expenditure was assessed by monitoring the CO2 production of cattle with the C-14-entry rate technique. Seven Angus steers (298 +/- 38 kg BW) were peritoneally infused with a solution of (NaHCO3)-C-14 for 72 hours using portable peristaltic pumps. The steers were forced to walk after 24 hours of infusion, on 2 consecutive days. On the first day, walking was at a constant speed of 2 km.hour(-1), divided in 4 periods of 0.5 hours (1 km), first on the level ground, second and third ascending and descending a 6% grade, and finally on the level surface on the way back to corrals. On the second day, cattle walked 1 km at 1 km hour(-1), and thereafter walked 4 km at 4 km hour(-1) on the level. Saliva samples were collected for periods of 0.5 hours before and during different periods of walking and at rest at 2 and 4 hours after the activity. Concentration and specific activity of CO2 were measured in saliva samples to estimate the rate of CO2 production [ml.hour(-1).(BW0.75)(-1)] as the ratio between the rate of infusion (mu Ci.hour(-1)) and the specific activity of CO2 (mu Ci.liter(-1) of CO2). The production of CO2 was converted to heat production using an energy equivalent of 536 kcal.liter(-1). Average energy expenditure (EE) in corrals in both days before the activity was 82.6 +/- 3.1 kcal hour(-1).100 kg BW-1 [650 mi CO2.hour(-1).(BW0.75)(-1)]. The cost of walking on the level surface and on the 6% grade was 9.0 +/- 1.14 and 16.4 +/- 2.18 kcal.km(-1).100 BW-1, respectively. There was a small nonsignificant residual effect of walking that disappeared a few hours after exercise. It was concluded that the cost of walking can only have a minor effect on the energy requirement of grazing cattle.