Abstract: This study investigates methodological discrepancies in particle size analysis to enhance the accuracy of soil gradation characterization. Comparative analyses were performed on eighteen distinct soil types,including aeolian sand,silty fine sand,sedimentary sand,loess,marine clay,and lacustrine clay,using two hydrometer techniques(Hydrometer A and B) and a laser particle size analyzer. The stability of these three methods was assessed. The correlation between test results was examined in relation to liquid limit characteristics and mineralogical composition. Furthermore,a sieve-based correction method was developed to refine the gradation results for sand and loess obtained from the laser particle size analyzer. The results indicate that: (1)The laser particle size method tends to underestimate sand particle content compared to the sieving method. The proposed correction can improve the accuracy of fine particle content measurement and soil classification; (2)A strong linear correlation between the laser particle size analyzer and hydrometer methods for silt content determination was observed,contrasted with weaker associations noted for sand and clay fractions; (3)A method stability hierarchy was established as laser particle size analyzer>Hydrometer A>Hydrometer B through statistical analysis,and hydrometer selection criteria were proposed based on liquid limit variations in loess samples; (4)The differences between the laser particle size analyzer and the hydrometer method are influenced by both the mineral density differentials and the clay mineral composition.