This study aimed to evaluate boron (B) adsorption and the capacity of a surface complexation model for simulating this process in calcareous soils. Ten surface soils were collected from different land use areas in Hamedan, Western Iran to characterize B sorption by soils. The mean B adsorbed by the sample soils varied from 8.9-32.8%. Two empirical models including linear and Freundlich equations fitted well to the experimental data. The linear distribution (Kd) values varied from 1.32 to 6.86 L kg-1, while the parameters of Freundlich equation including n and KFr ranged from 1.164-1.332 and 3.31-16.81, respectively. The comparison of two empirical models indicated that B adsorption followed a non-linear pattern. The soil organic matter (SOM) had positive correlations with Freundlich and linear distribution coefficients. However, empirical models were not suitable for explaining the mechanism of B adsorption, so a surface complexation model was used to simulate and predict the B adsorption process. B adsorption modeling was conducted using Visual MINTEQ and PHREEQC, based on the assemblage of major surface components (hydrous ferric oxides, aluminum hydroxides, calcium carbonate, and humic acids). B adsorption was successfully modeled by surface complexation. The significant contribution of organic matter to B complexes was resulted from both experimental data and mechanistic modeling.