Cycling becomes a more and more popular sport activity in the modern world. People want to reach their physical limits in training and competitions. Different kinds of measurements exist to give cyclists information about their ride (e.g. speed, torque) in real-time.This work deals with the possibility to use surface electromyography (sEMG) to infer muscle fatigue, cadence and symmetric usage of muscle groups during bicycling. Various methods have been used to extract low-level features from the EMG data. The features are energy, squared energy, mean frequency and energy in specic frequency bands. These features are subsequently used for statistical inference. The proposed technology allows for a system to be realized on a portable digital device (such as a smart-phone) which is either carried by the user or mounted on the bicycle. In a final product EMG signals could be given in real-time, e.g. by wireless communication with a sensor array that is strapped on the users legs. Fourteen ergometer bicycle rides from eight dierent subjectswere recorded and analyzed. The results show a strong relation between the sum of energy during the bicycle ride and the muscle fatigue. Frequency based features showed a low relation. The optimum feature set shows an acceptable predictor output for muscle fatigue assessment. The frequency analysis of the EMG signal allows a estimation of the cadence.