Objective:To evaluate the response inhibitory function of boxers with repetitive subconcussions using event-related potentials (ERPs) in the Go/Nogo paradigm.Methods:Totally, 29 boxers from Wuhan Physical Education Institute were enrolled in subconcussion group, and 28 college students from Wuhan Qingchuan University served as healthy control group. A 64-channel electrode elastic cap was used to record continuous scalp EEG data while completing the visual Go/Nogo experimental paradigm. E-Prime 2.0 software was used to record behavioral data. ERP components were extracted from EEG data, and the amplitude and latency of P3 component in ERP of the two groups were analyzed by repeated measure ANOVA. Two independent sample T tests were performed for Nogo-N2 component and behavioral data.Results:In ERP results, the mean amplitude of P3 in subconcussion group was lower than that in healthy control group under Go stimulation [(6.327±0.497) μV vs. (8.007±0.505) μV, F=5.622, P<0.05], the mean amplitude of P3 in subconcussion group was lower than that in healthy control group under Nogo stimulation [(6.741±0.559) μV vs. (9.962±0.569) μV, F=16.297, P<0.01], the latency of N2 in the subconcussion group was longer than that in the healthy control group under Nogo stimulation [(275.714±50.943) ms vs. (252.571±33.201) ms, t=-2.014, P<0.05]. In behavioral performance, the accuracy rate of subconcussion group was lower than that of healthy control group under Go stimulation [(97.3±1.9)% vs. (98.3±1.1)%, t=-2.476, P<0.05], and the accuracy rate of subconcussion group was lower than that of healthy control group under Nogo stimulation [(90.0± 6.6)% vs. (92.8± 3.3)%, t=-2.036, P<0.05]. Conclusion:These results preliminarily reveal that boxers with repetitive subconcussion may have cognitive impairments related to conflict detection and response inhibitory.