Résumé
This study aimed to assess the accuracy, reproducibility, and sensitivity of the Favero Assioma Duo (FAd) pedal-based power meter compared with the scientific-grade SRM (Schoberer Rad Messtechnik) crankset across a wide range of cycling intensities, from very low to maximal efforts. Eleven trained athletes performed submaximal and maximal cycling tests (10 '' to 20 ') on a home trainer while power and cadence data were simultaneously recorded from the FAd and SRM. Data were synchronized and analyzed to determine systematic bias, mean absolute error, and correlation coefficients under various conditions. Below 100 W, the FAd showed substantial relative errors, with deviations exceeding 200%. From 100 to 1200 W, the FAd slightly overestimated power (+1.2%) and cadence (+0.6%), with strong correlation to the SRM (r =.98). Maximal efforts revealed increasing error and decreasing r values as duration decreased, with the largest mean absolute error observed during the 10-second sprint (42.5 W). The FAd demonstrated acceptable reproducibility (coefficient of variation = 7.9%) and was sensitive to power differences as small as 3 W. Although the FAd lacks precision at very low intensities, it provides valid and consistent measurements for most training and performance contexts, including maximal efforts. Its accuracy, reproducibility, and sensitivity make it suitable for monitoring power output in both amateur and competitive cycling.