To accurately and comprehensively explore the entire process of physical fatigue development in rescue team members during weighted walking, a multidimensional fatigue assessment method based on eye movement characteristics, electromyographic signals, and subjective evaluation is proposed. Eight volunteers were recruited for the weight-bearing walking fatigue induction experiment. The glasses eye tracking was used to extract the eye movement data about ST (saccade time), average SS (saccade speed) and maximum SA (saccade amplitude). The correlation between the characteristics of eye movement and the degree of fatigue estimated by subjective evaluation was -0.857 ± 0.059, -0.938 ± 0.092, not correlated, respectively. The correlation with iEMG to judge fatigue degree was -0.782 ± 0.090, -0.942 ± 0.030, -0.928 ± 0.026, respectively. Multiple linear regression analysis was performed on subjective score, iEMG value and eye movement parameters. The regression model yielded a coefficient of determination R² = 0.989, with the following standardized coefficients: iEMG signals = 0.27, ST = -0.16, SS = -0.513, and SA = -0.124. This study makes new explorations and attempts in the monitoring and evaluation methods of fatigue during weighted walking.