[Objective] It is critical to investigate the ecological stoichiometry of soil nutrients and their driving mechanisms in wetlands , which providing key insights into soil nutrient cycling processes and internal driving factors. [Methods] This study focused on the wetlands of Banghu (semi-controlled lake) and Sizhoutou (open-water area) in the National Nature Reserve of Jiangxi Poyang Lake . Based on correlation analysis and structural equation modeling (SEM), we analyzed the ecological stoichiometric characteristics of surface soil carbon (C), nitrogen (N), and phosphorus (P) and their key influencing factors under different elevation gradients from 2017 to 2019. [Results] The mean concentrations of soil organic carbon(SOC), total nitrogen(TN), and total phosphorus(TP) in the Banghu wetland (9.33, 1.18 and 0.43 g/kg, respectively) were significantly higher than those in the Sizhoutou wetland (2.65, 0.31 and 0.19 g/kg, respectively) (p<0.05). At the same elevation gradient, SOC, TN, and TP concentrations in both wetlands did not vary significantly over the years (p>0.05). Although the mean soil C∶N ratio in the Banghu wetland (9.24) was significantly lower than that in the Sizhoutou wetland (10.19), the mean C∶P and N∶P ratios in the Banghu wetland (53.18 and 6.23, respectively) were significantly higher than those in the Sizhoutou wetland (40.90 and 4.05, respectively) (p<0.05). Elevation gradient had no significant effect on the C∶N ratio in either wetland (p>0.05). However, C∶P and N∶P ratios exhibited an increasing–decreasing trend along the elevation gradient. At the same elevation, C∶P in the Banghu wetland showed significant interannual variation (p<0.05), while no significant changes were observed in the Sizhoutou wetland (p>0.05). Both wetlands exhibited significant positive correlations of C∶P and N∶P ratios with aboveground biomass (AGB) and electrical conductivity (EC) and significant negative correlations with pH (p<0.05). Soil alkaline nitrogen (AN), ammonium nitrogen (NH4+-N), and AGB had direct effects on the soil ecological stoichiometry of both wetlands. In addition, the ecological stoichiometric characteristics in the Banghu wetland were directly influenced by soil bulk density (BD) and pH, while those in the Sizhoutou wetland were influenced by soil volumetric water content (SWC) and EC. In the Banghu wetland, AGB was the key factor influencing C∶N, while AN was the primary determinant of C∶P and N∶P. In the Sizhoutou wetland, NH4+-N was the main controlling factor for C∶N, and AGB was the critical factor affecting C∶P and N∶P. [Conclusion] Vegetation biomass and soil nitrogen play crucial roles in the ecological stoichiometric characteristics of soil carbon, nitrogen, and phosphorus in the Poyang Lake wetlands. The soil stoichiometric values in the semi-controlled Banghu wetland are primarily driven by AGB and AN, whereas those in the open-water Sizhoutou wetland are mainly influenced by NH4+-N and AGB. Therefore, techical model of vegetation restoration and reconstruction in wetlands could be considered to promote the ecological restoration and protection of Poyang Lake wetland in the future.