Argentina is the third largest producer of grain soybean in the world and the largest producer of soybean meal and soybean oil exports. Soybean is grown mostly in the Pampas of central–eastern Argentina. Argillic soils (typic and vertic argiudolls) and typic pelluderts are present in approximately 50% of the soils used for soybean production. Those soils have swelling and cracking behavior that may result in root clumping, and thus the root system’s ability to extract water may be restricted due to limited proliferation into the surrounding soil. The CROPGRO–Soybean model’s root growth and root water uptake routines were modified to improve its ability to simulate root growth and water uptake in soils where root penetration is restricted or roots are clumped together. Two empirical soil parameters, referred to as a soil impedance factor and a root clumping factor, were incorporated into this “revised” model. Data sets from a vertic argiudoll (Oliveros, Santa Fe), a typic argiudoll (Balcarce, Buenos Aires), and an enthic hapludoll (Manfredi, Cordoba) were used to evaluate the ability of the model to predict soil water extraction by soybean crops. An adaptive simulated annealing technique was used to estimate the soil impedance and root clumping factors. The revised CROPGRO–Soybean model was able to predict soil water extraction for three Argentine soils with different soil textures. The effectiveness of the model was expressed by low percent errors of estimation, ranging from 3.5% to 7.0% among different sites. The incorporation of soil impedance and root clumping factors allowed us to assess the effects of the argillic horizons in restricting root depth growth and root proliferation and thus in decreasing the potential root water uptake rates. The argillic horizons had more significant effects on root clumping than on restrictions of root depth growth. The clumping of roots caused by the presence of cracking horizons also reduced the rate of water uptake below those layers. Critical clay content thresholds as predictors of soil impedance and root clumping were found, providing a method for predicting soil impedance and root clumping factors for argillic soils.