Objective To explore whether there are short-term electrocardiogram characteristics which may serve as a biomarker for cumulative fatigue. Methods The electrocardiogram signals of 14 students (7 males and 7 females, with age ranging 19-23 years) from the University of Shanghai for Science and Technology, who were participating in the national electronic design competition for college students, were dynamically tracked. The scores of subjective fatigue scale and sleep duration were collected, and 13 characteristic electrocardiogram items were calculated, including heart rate variability, electrocardiogram entropy, high and low frequency values, and average RR interval. The difference between the electrocardiogram items was compared using the Wilcoxon signed-rank test between awakening and fatigue states. Results The subjective fatigue level was increased during lack of sleep, and the high frequency value of short-term electrocardiogram characteristics, the ratio of high to low frequency values and the Poincaré plot SD2 were significantly different between awakening and fatigue states (all P＜0.05). Conclusion Some short-term electrocardiogram characteristics could be used for the detection of cumulative fatigue.