Objective: To explore the eye movement characteristics of patients with cerebral small vessel disease (CSVD) and their correlation with cognitive impairment, and to validate the application value of eye-tracking technology in the early screening of CSVD-related cognitive impairment. Methods: A total of 81 subjects were included, with 51 in the CSVD group and 30 in the healthy control group. Cognitive assessment was conducted using the MoCA scale, and visual pairing and antisaccade tasks were detected using an AI-assisted eye-tracking system. Differences in eye movement parameters between the two groups were compared, and the correlation between eye movement indicators and cognitive scores was analyzed. Results: The saccade error rate in the CSVD group was significantly higher than that in the control group (P = 0.009). The total MoCA score, visuospatial and executive function, attention, and language function scores were significantly lower in the CSVD group than in the control group (all P < 0.05). The saccade error rate was significantly negatively correlated with the total MoCA score (rs = -0.382, P = 0.0005) and visuospatial and executive function score (rs = -0.313, P = 0.005). Conclusion: Patients with CSVD exhibit characteristic eye movement abnormalities, and the saccade error rate can serve as an objective biomarker reflecting their executive function and overall cognitive impairment. Eye-tracking technology provides a non-invasive, objective, and efficient new means for the early screening of CSVD-related cognitive impairment.