Hexavalent Chromium (Cr(VI)) contamination in soils poses significant ecological risks due to its mobility and toxicity, with retention mechanisms governed by interactions between soil properties and Cr(VI). However, the quantitative roles of key soil parameters in Cr(VI) retention remain poorly resolved, particularly across diverse soil types. This study investigated Cr(VI) retention behaviors in 16 Chinese soils (15 types) through batch experiments, isothermal adsorption model, correlation analysis and path analysis. The results showed that the retention of Cr(VI) in acidic soils was significantly higher than in alkaline soils. Acidic soils (pH < 5.4) with higher concentrations of exchangeable Fe(II) (Exch-Fe(II)) exhibited strong Cr(VI) holding capabilities,while Alkaline soils (pH > 7.3) with highest content of CaCO ₃ show negligible Cr(VI) reactions.Cr(VI) retention was high at soil pH values below approximately 5.5, but declined sharply at higher pH values. The Langmuir model was only suitable for describing acidic soils (pH < 5.4), while the Freundlich equation was applicable to all soils. Correlation analysis revealed that soil pH, the content of soil organic matters(SOM), Exch-Fe(II), complexed iron (Com-Fe), and clay were significantly related to the Cr(VI) retention ( p < 0.01), whereas the CaCO ₃ content was negatively related to the Cr(VI) retention ( p < 0.05).Path analysis revealed that soil pH was the most important direct factor, followed by Exch-Fe(II), Com-Fe, clay, in determining Cr(VI) retention in natural soil. CEC and CaCO ₃ content had only limited directly effects on the Cr(VI) retention. Additionally, The content of SOM, Amorphous iron oxides(Amo-Fe), and Easily reducible manganese(Er-Mn) content had little directly effect on Cr(VI) retention. To validate these findings, Cr(VI) retention was measured in all soils after adjusting their pH to 4.3, 6, and 8. The results highlighted soil pH and Exch-Fe(II) content were the most decisive factors for evaluating Cr(VI) retention in natrual soils,whereas SOM content was an unreliable parameter for assessing this process.