Argentine agricultural production is fundamentally based on a technological package that combines direct seeding and glyphosate with transgenic crops (soybean, maize and cotton), which makes glyphosate the most widely employed herbicide in the country. Glyphosate is strongly sorbed to soil in a reversible process that regulates the half-life and mobility of the herbicide, with the resulting risk of contaminating surface and groundwater courses. However, this behavior may vary depending on the characteristics of the soil on which it is applied. Sorption coefficients are thus the most sensitive parameters in models used for environmental risk assessment. The aim of this work was to study the affinity of glyphosate to 12 different soils of Argentina and create a model to estimate the glyphosate Freundlich sorption coefficient (Kf) from easily measurable soil properties. Batch equilibration adsorption data are shown by Freundlich adsorption isotherms. Principal component analysis and multiple linear regressions were used to correlate the effects of soil properties on glyphosate adsorption coefficients. Results indicate that pH and clay contents were the major soil parameters governing glyphosate adsorption in soils. The Freundlich (Kf) pedotransfer function obtained by stepwise regression analysis has 97.9% of the variation in glyphosate sorption coefficients that could be attributed to the variation of the soil clay contents, pH, PBray and Alin. Display omitted •We used multiple linear regression to predict glyphosate Kf from soil properties.•PCA was used to correlate the effects of soil properties on Kf.•Glyphosate sorption appeared to be mainly controlled by pH and clay content.•Four key soil parameters provide a robust pedotransfer function for Kf prediction