OBJECTIVE To investigate the numerical relationship between the changes of surface electromyographic (sEMG) signal indexes and subjective assessment of muscle fatigue during isometric contractions. METHOD Eleven healthy subjects participated in the experiment. The sEMG signals of biceps were recorded during their sustained isometric contractions at 50% maximum voluntary contraction (MVC). Fatigue was subjectively rated using a Borg CR-10 scale. Correlations between Borg ratings and a series of fatigue-related indexes such as MPF, MF and Lempel-ziv complexity were analyzed, and curve estimation was also explored. RESULT A strong positive correlation was found between Borg ratings and endurance time, and negative correlations at different degrees were found between Borg ratings and each of sEMG signal indexes. Curve estimation (Logarithmic curve and Power curve) was also significant. CONCLUSION The significant correlation between the Borg scale, sEMG signal indexes and endurance time indicates a close relationship between subjective and objective assessment of muscle fatigue. The numerical relationship between them accord with the power law of Stevens in psychophysics relatively.