Per- and polyfluoroalkyl substances (PFAS) have been linked to kidney function. Studies have shown that PFAS can cause changes in lipid metabolism and that lipids play an important role in regulating kidney function. However, few studies have explored the overall impact of PFAS mixture on kidney function. Moreover, the mechanisms by which PFAS influences kidney function remain unclear. This study was performed to investigate the overall impact of PFAS mixture on kidney function indexes, dissect the mechanism by which PFAS affect kidney function by analyzing lipid molecule profiles, and analyze the associations between different subclasses of lipids and kidney function indexes. We measured blood PFAS levels and kidney function indexes in a community population containing 278 males. Metabolomic analysis detected 332 lipid molecules. A quantile-based g-computation model was applied to assess the overall effect of PFAS mixture on kidney function index, and revealed that PFAS mixture were associated with a higher level of uric acid (UA). Linear regression analysis demonstrated a positive association between PFOA and UA, and logistic regression analysis indicated a positive association between PFOA and hyperuricemia odds. Notably, none of the PFAS were associated with the estimated glomerular filtration rate, indicating that PFAS didn't have an obvious effect on glomerular filtration. Further analysis identified 20 lipid molecules associated with both PFOA and UA. High-dimensional mediation effect analysis showed that seven lipid molecules (one glycerophospholipids, three fatty acyls and three prenol lipids) mediated the association between PFOA and UA. Additionally, quantile-based g-computation analysis revealed positive associations between specific lipid subclasses-mainly fatty acid esters, fatty acids and conjugates, and sesquiterpenoids-and kidney function indexes. Our findings provide insights into the renal toxicity of PFAS and may also lead to more in-depth investigations using animal models and other population studies.