The current laboratory adsorption study aimed at determination of the values of adsorption distribution coefficient (Kd) of bentazone in the profiles of Arenosols, Luvisols, and Cambisols, which are the most common arable mineral soils in Poland. The study attempted to identify the soil components that bind bentazone and the principal adsorption mechanisms of this compound as well as create a model capable of predicting its adsorption in soils. The Kd values determined in batch experiments after 24 h of shaking were very low, and ranged from 0.05 to 0.30 mL/g for the Ap horizon and 0 to 0.07 mL/g for subsoils. The results indicated that the anionic form of bentazone was adsorbed on organic matter, while in acidic soils the neutral form of bentazone was adsorbed on organic matter and sand. The detailed analyses of mineralogical composition revealed that the principal mineral that was responsible for the adsorption of bentazone was quartz, which content was strongly positively correlated with the sand fraction. In soils with pH < 5 and an organic carbon content of < 0.35%, quartz exhibited much greater affinity for the neutral bentazone form than organic matter. Fourier transform infrared photoacoustic spectroscopy analyses supported by computational methods have shown the most probable mechanisms behind the adsorption of bentazone on quartz. The created model, assuming the adsorption of bentazone on organic matter and on sand and using the spectrophotometrically determined dissociation constant of bentazone, very well explained the Kd variance in the 81 examined soils, while correctly predicting the adsorption based on soil properties described in the published data.