The sorption of Sb(V) onto natural calcareous soils was studied in batch and column experiments as a function of physicochemical properties of the soil, namely: organic matter, the active fraction of Fe and Al and the pH of the soils. Batch experiments were performed in order to determine the sorption capacity and the kinetic rate of the sorption. Freundlich isotherm described properly the equilibrium experimental data and the kinetic results show that the sorption was very slow in all the soils studied (equilibrium after 7 days) when compared with literature data for Sb(V) sorption onto soils with higher iron content. Column experiments were carried out in order to determine the mobility of Sb(V) in the soils. The Thomas model can describe partially the breakthrough curves obtained for the Sb(V) sorption on the soils. The fitting of the model to the experimental data would indicate that transport and kinetic parameters also affected the dynamic sorption behavior. The experimental breakthrough curves can be properly described by the convective–dispersive equation (CDE) by using the CXTFIT code, and the parameters obtained by the model were in concordance to the sorption capacity reported in the column experiments. This study indicates that sorption of Sb(V) onto the soils is mainly affected in batch experiments by their ‘active’ Fe and Al content; while in column experiments organic matter content appeared as an important parameter. Desorption experiments resulted in a metalloid recovery above 90%, which indicates a fast and effective method to recover Sb(V) from calcareous soils and has implications for its geochemical behavior (vertical mobility in soil profiles) as a result of its relatively low Sb retention.