Indirect calorimetry is used routinely to estimate energy expenditure in athletes. By means stoichiometric equations, carbohydrates (CHOOXR) and fat (FATOXR) are estimated from VO2 and VCO2 (1). This methodology is useful to understand exercise-induced metabolic adaptations and athletes’ fuel needs. There are different stoichiometry equations in literature (2), but no studies had evaluated the validity of all these equations respect to a gold standard for whole intensities of aerobic exercise. The present study was designed to assess the agreement between the different stoichiometric equations vs. the breath 13C/12C ratio technique as a gold standard. 16 male endurance well-trained athletes (n=1 Olympic level in Triathlon) (BMI: 22.3±6.2kg/m2, VO2Peak: 4.05±0.50L/min) followed a glycogen depletion-repletion protocol to increase the 13C enrichment of endogenous CHO (3). The day after, subjects performed a laboratory test on cycloergometer: 10min warm-up at 2.0W/kg and increments of 0.5W/kg/10min until exhaustion. VO2 and VCO2 were measured during the test (Oxycon-Pro®). CHOOXR and FATOXR were estimated for all different equations (2) at resting and at the end of stages. Isotope infusion, 13C plasma enrichment of CHO, fats, and proteins were measured; as well as CHOOXR and FATOXR from 13C/12C breathing (3). The agreement between estequiometric equations vs. the gold standard was assessed by Bland-Altman (4). In resting (6-16%VO2Peak) all FATOXR equations had the same agreement (-0.03±0.18g/min) and all CHOOXR equation showed similar agreements being the Brouwer the best one (0.08±0.46g/min). All FATOXR equations presented good agreements at low-moderate intensities of exercise (19-80%VO2Peak), to moderate-high intensities (81-100%VO2Peak) all FATOXR equations presented moderate errors, being the lowest one for the Brouwer’s equation (Lower limit agreement (LLA): -0.46 to 0.10, Bias: -0.22 to 0.26 and Upper limit agreement (ULA): 0.02 to 0.43 (g/min)). For CHOOXR equations at low-moderate intensities the Frayn’s and the Ferranini’s (for glucose) equations showed the lowest error (LLA: -0.09 to 0.01, Bias: -0.01 to 0.06 and ULA: -0.02 to 0.21 (g/min)), and the Ferranini (for Glycogen) one at moderate-high intensities (LLA: -0.96 to -0.20, Bias: -0.48 to 0.77 and ULA: 0.00 to 1.75 (g/min)). In summary, from resting to medium intensities of exercise all FATOXR equations can be used indistinctly but at moderate-high intensities the Brouwer one is the most suitable to use. The best CHOOXR equations were the Brouwer one for resting, the Frayn and the Ferranini (for Glucose) ones at low-moderate intensities, and the Ferranini (for Glycogen) one at moderate-high intensity.