Objective: Walking is an important mode of exercise and is likely to represent a major component of nonexercise activity thermogenesis. The question arises, how best to quantify walking-energy expenditure (EE) in free-living individuals. The triaxial accelerometer for movement registration (Tracmor) is a valid measure of body displacement and so we wanted to evaluate this tool for quantifying walking-EE. Hypothesis: In this study, we test the hypothesis that walking-EE, measured in a Room Calorimeter, can be predicted from Tracmor output using a regression equation derived from a brief Tracmor/treadmill/Metabolic Cart protocol. Design: First, 11 healthy subjects completed a 40-min procedure whereby they wore a Tracmor unit and walked on a treadmill at 0, 1, 2 and 3 mph while EE was measured using a Metabolic Cart. This allowed a regression equation to be defined for each subject to convert Tracmor output to EE. Each subject then entered a Room Calorimeter wearing the Tracmor Unit and walked at two self-selected velocities ('slow', 'fast') while EE was measured. 'Tracmor/regression equation' predictions of walking-EE were compared with Room Calorimeter measurements of walking-EE for the two velocities. Results: The 'Tracmor/regression equation' prediction of EE for walking slowly was 6.36±1.67 kJ/min, and for walking fast it was 11.0±2.60 kJ/min. Room Calorimeter measurements were 6.43±1.85 and 10.9±3.03 kJ/min, respectively. The intraclass correlation coefficient for slow-paced walking was 0.93 (P<0.001), and for fast-paced walking it was 0.82 (P<0.005). Conclusions: When combined with laboratory measures of EE, the Tracmor accelerometer provides useful data on walking-EE and is applicable to free-living individuals.