Measurement of energy expenditure in free-living individuals represents a methodologic challenge in epidemiologic research. Heart rate monitors, while closely tied to energy expenditure at high levels of energy output, provide much less predictive power at low levels; however, measurement of motion may improve the predictive ability. This study was undertaken to determine the usefulness of simultaneously monitoring heart rate and motion for the estimation of energy expenditure. Ten subjects were studied during simulated activities of daily living (ADLC) and submaximal treadmill tests. Compared to direct measurement, the motion sensor predicted oxygen consumption poorly (r2 = 0.53) for both tests. Heart rate measured simultaneously yielded an r2 of 0.81 for ADLC and 0.90 for the treadmill. Addition of motion data increased the r2 value for the ADLC for all but one individual and increased the group mean from 0.81 to 0.86. This improvement was not observed for the treadmill, confirming the hypothesis that the principle value of monitoring motion occurs at lower heart rates.