Objective Predictive equations of resting metabolic rate(RMR)are important tools for control of energy balance in nutrition support and the prevention and treatment of obesity.However，the best one for non-obese healthy Chinese adults is unclear.Our objective is to examine the accuracy of 14 common RMR predictive equations and to find the best alternative for indirect calorimetry.Methods RMR was measured by indirect calorimetry and calculated by 14 predictive equations in 46 participants.Bias，accuracy rate，root mean squared prediction error and concordance correlation coefficient were used to evaluate accuracy of the predictive equations.Results Although the biases of H-B，Mifflin，WHO，Schofield，Livingston，Henry，Muller，Lazzer，De Lorenzo and Huang equations were small(<5%)，they were less accurate in the individual level.All of the accuracy rates were no more than 70%.The Henry equation(69.6%)had the highest accuracy rate，followed by Lazzer(67.4%)，H-B(67.4%)，Mifflin(65.2%)，Schofield(65.2%)，WHO(65.2%)，De Lorenzo(65.2%)，Muller(63.0%)and Livingston(60.9%). Bernstein equation was the most inaccurate.Conclusions All of the 14 common predictive equations were not accurate enough in non-obese healthy Chinese adults.Henry equation was relatively the most accurate and may act as the best alternative for indirect calorimetry when indirect calorimetry was unavailable.