OBJECTIVE: The objective of this large sample cross-sectional study was to investigate the relationships between body roundness index (BRI), atherogenic index of plasma (AIP), and cognitive impairment in older Chinese people. METHODS: This study included a population of 97,081 individuals who underwent health check-up at People's Hospital of Jingning She Autonomous County from 2020 to 2024. Cognitive function was assessed using the Montreal Cognitive Assessment (MoCA). Data was collected through face-to-face interviews, physical examinations, and laboratory tests. BRI was calculated and grouped into quartiles. Univariate logistic regression models and multivariate logistic regression models were used for analysis. The restricted cubic spline (RCS) curve was used to investigate the dose-response relationship between BRI levels and cognitive impairment, and a mediation analysis was conducted to examine the role of AIP in mediating the effect of BRI on cognitive impairment. RESULTS: The likelihood of cognitive impairment increased by 67.1% for each unit increase in AIP (OR = 1.671, 95%CI: 1.258-2.221, p < 0.001), and participants in the third quartile of BRI (Q3) had a 31.2% lower risk of cognitive impairment compared to those in the first quartile (Q1) (OR = 0.688, 95%CI: 0.528-0.897, p = 0.006). There was a nonlinear U-shaped relationship between BRI levels and cognitive impairment (P for nonlinear = 0.042), although the p value for overall was not statistically significant. Furthermore, AIP statistically accounted for the association between BRI and cognitive impairment (p < 0.001). CONCLUSION: Elevated AIP was independently associated with a higher risk of cognitive impairment in older Chinese adults, and statistically accounted for a substantial portion of the BRI-cognitive impairment association. The protective signal at intermediate BRI levels, without a linear trend, likely reflects a localized phenomenon rather than a global obesity paradox. Longitudinal studies are needed to confirm directionality.