Abstract: Monitoring the dynamic variations in water levels of lakes and reservoirs in the middle and lower reaches of the Yangtze River is crucial for understanding the regional hydrological cycle mechanisms. In this study, 12 lakes and reservoirs of different sizes within the region were selected for analysis. Water level time series were first extracted from SWOT data products at different spatial resolutions along with the corresponding water level quality evaluation indices (wse_qual). The water levels derived from SWOT data were then systematically compared with water level datasets from DAHITI, Hydroweb, and available in situ observations. To quantitatively evaluate the accuracy of SWOT water level measurements, statistical metrics including the root mean square error (RMSE), correlation coefficient (R), and mean absolute error (MAE) were employed. The results indicate that the RMSE, R, and MAE between SWOT water levels at a spatial resolution of 100 m and DAHITI water levels for eight lakes and reservoirs range from 0.1593 to 0.511 m, 0.7557 to 0.988, and 0.1217 to 0.4656 m, respectively. For five lakes and reservoirs with available in situ observations, the corresponding RMSE, R, and MAE between SWOT water levels and observed water levels range from 0.1122 to 0.2277 m, 0.9538 to 0.9988, and 0.0714 to 0.1666 m, respectively, indicating strong correlation and consistency, with an overall accuracy at the decimeter level. Moreover, compared with DAHITI, Hydroweb water levels exhibit even higher agreement and coherence with SWOT observations. the SWOT data product at a spatial resolution of 100 m demonstrates greater stability and reliability in monitoring lake and reservoir water levels under different wse_qual conditions, maintaining high accuracy even for small water bodies; In contrast, the accuracy of the data products with a spatial resolution of 250 m shows notable discrepancies. For example, when the wse_qual value changes from 0 to 0/1, the RMSE between SWOT and DAHITI water levels at Zhelin Reservoir increases from 0.4994 m to 0.7210 m.Overall, these results validate the capability of the SWOT satellite to accurately monitor water levels of lakes and reservoirs across a wide range of sizes, providing a robust data foundation for high-precision regional hydrological monitoring and water resource assessment.