The Sodium Adsorption Ratio (SAR) is a critical indicator for characterizing the hazard of sodium ions and the degree of soil sodification in saline-alkali soils. However, in the coastal region of Hebei, the characteristics and key influencing factors of soil SAR remain unclear due to unique processes of salt formation and accumulation, as well as complex physicochemical interactions, which hinders its accurate prediction.【Objective】This study aims to elucidate the spatial distribution and profile variation patterns of SAR in representative coastal saline-alkali soils of Cangzhou, Hebei; to identify and quantify the key soil physicochemical factors influencing SAR dynamics. Also, the study seeks to develop and select an optimal machine learning model for accurately predicting SAR based on easily measurable parameters.【Method】Taking typical coastal saline-alkali land in Cangzhou City, Hebei Province as the research area, soil samples were collected from two layers (0–20 cm and 20–40 cm). A comprehensive set of properties was measured, including ionic composition, bulk density (BD), soil water content (SWC), soil organic matter (SOM), total porosity (STP), capillary porosity (SCP), saturated hydraulic conductivity (Ks), electrical conductivity (EC), and pH. The characteristics of SAR were analyzed, its main influencing factors were explored, and four machine learning models: Linear Regression (LR), Decision Tree (DT), Random Forest (RF), and K-Nearest Neighbors (KNN), were used to predict SAR. Model performance was evaluated using the coefficient of determination (R2) and Root Mean Square Error (RMSE).【Result】The mean SAR values in the upper (0–20 cm) and lower (20–40 cm) layers were 22.23 and 28.02, respectively, with no significant difference (P = 0.126). The soil in the study area was classified as moderately saline-sodic soil. Correlation analysis revealed that SAR was significantly correlated with K?, Cl?, SO?2?, EC, pH, BD, HCO??, SOM, SWC, STP, SCP, and Ks. Among these, the correlations with Cl?, SO?2?, and EC were the strongest, identifying them as the primary influencing factors. In the comparison of SAR prediction models, a model using both EC and pH as predictors achieved higher accuracy, and the RF model demonstrated the best predictive performance, with soil EC being the most significant feature.【Conclusion】The RF model can achieve robust prediction of SAR in the coastal saline-alkali soils of Hebei based on easily measurable indicators such as EC and pH. This study identified the key driving factors of SAR in the region and developed an effective predictive framework, providing a scientific basis and practical tools for the precise reclamation and sustainable utilization of local saline-alkali lands.