【Objective】The Laser Diffraction Method(LDM) has been increasingly adopted for determining Soil Particle Size Distribution(PSD) due to its rapid measurement speed, high degree of automation, and minimal labor requirement. These advantages make LDM particularly attractive for large-scale soil surveys, laboratory routine analysis, and studies requiring high-resolution particle-size information. Nevertheless, substantial discrepancies persist between PSD measured by LDM and that obtained from the traditional Sieve-Pipette Method(SPM), which remains the reference method in most national and international soil texture classification systems. These inconsistencies arise mainly from differences in measurement principles, contrasting assumptions regarding particle shape and density, the optical models used in LDM instruments, and the absence of unified sample pretreatment protocols. As a result, the direct use of LDM-derived PSD for soil texture classification often yields biased or inconsistent outcomes. Therefore, this study aims to: 1) comprehensively compare PSD measured by LDM and SPM for major soil types in the Black Soil Region(BSR) of Northeast China; and 2) develop and evaluate calibration models based on regression analysis and Lin’s Concordance Correlation Coefficient(CCC) to convert LDM results into an SPM-compatible reference framework. 【Method】Soil samples representing 36 dominant soil series across the BSR, including black soil, chernozem, chestnut Soil, and aeolian Sandy Soil, were collected in the field. After air-drying, the samples were gently ground using a mortar and passed through a 2 mm sieve. PSD for each sample was determined using both LDM and SPM following their respective protocols. Differences in clay, silt, and sand fractions were quantified to characterize systematic deviations. Two calibration strategies were developed: one based on regression analysis, which establishes quantitative relationships between LDM and SPM values using linear regression models; and the other based on Lin’s CCC, which adjusts particle-size cutoff thresholds in LDM to maximize consistency with SPM and identify optimized boundary values for clay, silt, and sand. Model performance was evaluated based on reductions in mean absolute differences and improvements in soil texture classification accuracy using the USDA textural classification system. 【Result】1) Pronounced discrepancies were observed between LDM and SPM. Relative to SPM, LDM consistently underestimated clay content and substantially overestimated silt content across all soil types, whereas deviations in sand content varied depending on sample-specific characteristics. The mean absolute differences for clay, silt, and sand were -30.1%, 34.5%, and -4.4% respectively. 2) Following regression-based correction, the mean absolute differences decreased to 0.9%, 0.5%, and -1.4%, and the accuracy of soil texture classification improved markedly to 36.1%. 3) Lin’s CCC-based calibration further reduced the mean absolute differences to 1.3%, -3.4%, and 2.1%, yielding a classification accuracy of 38.9%. Although both calibration approaches substantially improved agreement between LDM and SPM, the regression-based method was more computationally efficient to apply in routine laboratory workflows. 【Conclusion】Uncorrected LDM-derived PSD is unsuitable for direct soil texture classification in the BSR. However, applying appropriate calibration models, whether regression-based or Lin’s CCC-based, effectively harmonizes LDM data with SPM reference values, significantly enhancing its reliability and practical utility. These findings offer a methodological foundation for integrating LDM into soil survey, monitoring, and classification programs, and provide valuable guidance for improving the comparability of PSD data obtained using different analytical techniques.