Muscle fatigue is exhibited in individuals who partake in repetitive movements for an extensive duration of time. The objective of this study was to identify correlation, if any, between movement kinematics and muscle electrical activity (EMG). Movement kinematics and EMG signals were measured and recorded from the lower extremities of stationary elite cyclists. Standard statistical metrics (SSM) and phase space warping (PSW) based features were extracted from the recorded fast-time time series for consecutive intermediate time intervals. The SSM based features were composed of higher moments, fractal dimension, correlation sum, power spectral density, and cross-correlation. The PSW based features were subject to quantifying shorttime differences between the fatigued and unfatigued reconstructed phase spaces. The slow-time manifolds describing global fatigue dynamics in these feature spaces were extracted using smooth orthogonal decomposition (SOD). Mean and median frequencies from the EMG data were calculated to describe the local fatigue dynamics in each muscle. There were very close correlations between the EMG and kinematics data based global fatigue features. Also, the 4 and 5 dimensional slow time manifolds (corresponding to PSW and SSM based features, respectively) fully represented the local fatigue dynamics in all the muscles as described by the EMG data. Therefore, for this particular context, the fatigue information present in the standard EMG analysis was fully represented in the SOD based slowtime features extracted from the kinematic data. Furthermore, the SOD based analysis gave estimates of effective dimensionality of muscle fatigue dynamics.