With the rapid development of global aquaculture, the welfare of farmed aquatic animals has become a growing concern. As an emerging environmental stressor, underwater noise pollution has garnered significant attention in ecotoxicological research due to its impacts on fish auditory systems and behavioral patterns. The large yellow croaker (Larimichthys crocea), a representative species of the Sciaenidae family, exhibits high auditory sensitivity. However, impulsive low-frequency noise (800–1200 Hz) generated by construction activities (e.g., engineering drilling) in coastal aquaculture zones overlaps significantly with the fish’s most sensitive auditory frequency range (400–600 Hz), potentially causing hearing impairment and behavioral stress. While existing studies have demonstrated that low-frequency acoustic stimuli affect physiological indicators in L. crocea, direct evidence of hearing damage and the regulatory effects of body size remain insufficiently explored. This study aims to address these gaps, providing data to optimize aquaculture environments, enhance fish welfare, and establish noise management standards. Juvenile large yellow croakers from the Rudong Institute in Jiangsu were subjected to short-term noise exposure experiments in a 6 m × 1.5 m × 1 m concrete tank. The noise source was an engineering drill (128T AVT HUMMER) operating 3–10 meters from the tank. Construction activities were conducted daily from 7:00 to 11:00 a.m. for 10–20 minutes per session, with 30-minute intervals, over three days. A Reson hydrophone (TC4032) and Brüel & Kj?r data acquisition module were used to record the sound pressure level (SPL) and particle motion (PM) of the construction noise. Spectral analysis revealed that the noise’s primary frequency range was 800–1200 Hz, with intensities 40–60 dB higher than the baseline noise level in the aquaculture tank (60–80 dB). Auditory evoked potential (AEP) experiments were conducted in a 50 cm diameter cylindrical tank using a UW-30 underwater speaker to deliver pure-tone stimuli (100–1200 Hz, 130–60 dB, 3 dB steps). AEP signals were recorded using a TDT auditory electrophysiology workstation. The experiment included two phases: pre-exposure (control group) and post-exposure (experimental group). Each fish was tested at 10 frequencies (100–1200 Hz) to determine auditory thresholds. Generalized linear mixed models（Generalized Linear Mixed Models, GLMMs） were employed to analyze the interaction effects of auditory thresholds with frequency, body weight, and noise exposure. Fixed effects included frequency, body weight, and group (control/experimental), while random effects accounted for individual variability. Post-exposure auditory thresholds in L. crocea increased significantly (p < 0.001), with the greatest threshold elevation observed in the most sensitive frequency range (400–600 Hz). At 500 Hz, the average hearing loss reached 6.36 dB, indicating substantial damage to critical communication frequencies. Low-frequency (100 Hz) and high-frequency (1000 Hz) thresholds stabilized at 99.75 dB (±1.97) and 113.5 dB (±2.35), respectively, while mid-frequency thresholds (200–800 Hz) showed greater variability, suggesting stronger noise-induced disruption in mid-range hearing. At 300 Hz and 400 Hz, body weight exhibited a significant positive correlation with auditory thresholds (Spearman r = 0.673, p = 0.033; r = 0.753, p = 0.012), indicating that larger individuals had reduced auditory sensitivity. The GLMMs model revealed a significant interaction between body weight and noise exposure (p = 0.004), with the positive effect of body weight on auditory thresholds being more pronounced in the experimental group (noise-exposed). At 700 Hz, a significant positive correlation (p = 0.013) was observed, suggesting greater high-frequency hearing loss in larger individuals. This study systematically evaluated the effects of construction noise exposure on the auditory thresholds of L. crocea and explored the regulatory role of body size in auditory sensitivity. By integrating auditory evoked potential (AEP) techniques and generalized linear mixed models (GLMMs), the research elucidated the damage characteristics and potential mechanisms of construction noise on the fish’s auditory system, providing a scientific basis for noise management in aquaculture environments. The overlap between the construction noise’s dominant frequency (840 Hz) and the fish’s most sensitive auditory range (400–600 Hz) resulted in the greatest hearing loss in critical communication frequencies, potentially disrupting acoustic communication and environmental perception. Larger individuals exhibited reduced auditory sensitivity in mid-frequency ranges (300–400 Hz), a pattern consistent with findings in other species (e.g., Scorpaenodes barbatus), likely linked to auditory system development and sound propagation efficiency. Noise-induced hearing impairment and stress behaviors may reduce foraging efficiency and reproductive success in L. crocea, ultimately affecting aquaculture yields. Furthermore, the observed cumulative effects of noise exposure highlight the need to address the long-term risks of chronic noise pollution. This study focused on juvenile fish; future research should extend to adults and reproductive-stage individuals to assess developmental sensitivity differences. Additionally, experiments conducted in closed tanks differ from open-sea net cage environments, necessitating validation in natural conditions. Long-term studies should incorporate histological analyses (e.g., inner ear hair cell damage) and behavioral ecological metrics to comprehensively evaluate the population-level impacts of noise.