An experiment was conducted to test the validity of a mechanistic simulation model describing the process of soil moisture extraction by root systems, and to assess the quantitative importance of root hydraulic resistance in this process. Close-growing stands of ryegrass were raised in tubes packed with sand and loam, with and without the presence of a water table. The pattern of water uptake was monitored by gamma-ray scanning and tensiometry. Correspondence between measured and independently predicted uptake patterns could be greatly improved by assigning a relatively high value of root resistivity. Further improvement of the mechanistic model's predictive capability may require explicit formulation of conductive (axial) resistance in addition to absorptive (radial) resistance, as well, perhaps, as such important subsidiary factors as aeration, mechanical impedance, and distribution of nutrients in the soil profile.