Abstract: Objective:The purpose of this study was to analyze the effectiveness of SWA sensors to predict the energy expenditure during running,walking and cycling,providing a reference for predicting energy consumption of sports and fitness.Methods:15adult males in the treadmill exercise make completion of 4.8km/h,6.4km/h,8.0km/h three speed of running and walking,and completed in the indoor 10km/h,13km/h,and 15km/h cycling.The energy consumption was predicted by SWA and indirect calorimetry.Statistical analysis included paired t test,regression analysis,linear correlation analysis,ICC coefficient and Bland-Altman method for the forecast data.Results:1)SWA6.1overestimated the energy consumption of treadmill running and walking(1%)and field-cycling(9%),but the relevance and consistency(r= 0.91,ICC=0.91)were good.2)SWA7.0overestimated the energy consumption of treadmill running and walking(over 325%)and field-cycling(over 357%),but the relevance were good(r=0.91),and the consistency was acceptable(ICC=0.44).The consistency of SWA7.0and IC in predicting running and walking energy consumption was greater than that when SWA7.0predicted cycling energy consumption.Conclusion:SWA sensor highly overestimated the energy consumption of sports and fitness,and the overestimated value is higher and high with increasing exercise intensity,with the previous version of the SWA sensor overestimated the energy consumption of low intensity sports and underestimated the high intensity sports,but there were high relevance and consistency between the predicting energy consumption of SWA sensor and IC.