摘要:Electromyography (EMG) is normalized in relation to a reference maximum voluntary contraction (MVC) value. Different normalization techniques are available but the most reliable method for cycling movements is unknown. This study investigated the reliability of different normalization techniques for cycling analyses. Twenty-five male cyclists (age 24.13 ± 2.79 years, body height 176.22 ± 4.87 cm and body mass 67.23 ± 4.19 kg, BMI = 21.70 ± 2.60 kg·m−1) performed different normalization procedures on two occasions, within the same testing session. The rectus femoris, biceps femoris, gastrocnemius and tibialis anterior muscles were examined. Participants performed isometric normalizations (IMVC) using an isokinetic dynamometer. Five minutes of submaximal cycling (180 W) were also undertaken, allowing the mean (DMA) and peak (PDA) activation from each muscle to serve as reference values. Finally, a 10 s cycling sprint (MxDA) trial was undertaken and the highest activation from each muscle was used as the reference value. Differences between reference EMG amplitude, as a function of normalization technique and time, were examined using repeated measures ANOVAs. The testretest reliability of each technique was also examined using linear regression, intraclass correlations and Cronbach’s alpha. The results showed that EMG amplitude differed significantly between normalization techniques for all muscles, with the IMVC and MxDA methods demonstrating the highest amplitudes. The highest levels of reliability were observed for the PDA technique for all muscles; therefore, our results support the utilization of this method for cycling analyses.