Root systems of individual crop plants may encounter large variations in mechanical impedance to root penetration. Split-root experiments were conducted to compare the effects of spatial variation in soil strength on the morphological plasticity of wheat and barley roots, and its relationship to shoot growth. Plants of spring barley (Hordeum vulgare cvPrisma) and spring wheat (Triticum aestivum cv Alexandria) were grown for 12 days with their seminal roots divided between two halves of a cylinder packed with sandy loam soil. Three treatment combinations were imposed: loose soil where both halves of the cylinder were packed to 1.1 g cm-3 (penetrometer resistance 0.3 MPa), dense soil where both halves were packed to 1.4 g cm-3 (penetrometer resistance 1 MPa), and a split-root treatment where one half was packed to 1.1 and the other to 1.4 g cm-3. In barley, uniform high soil strength restricted the extension of main seminal root axes more than laterals. In the split-root treatment, the length of laterals and the dry weight of main axes and laterals were increased in the loose soil half and reduced in the dense soil half compared with their respective loose and dense-soil controls. No such compensatory adjustments between main axis and laterals and between individual seminal roots were found in wheat. Variation in soil strength had no effect on the density of lateral roots (number per unit main axis length) in either barley or wheat. The nature and extent of wheat root plasticity in response to variation in soil strength was very different from that in response to changes in N-supply in previous experiments. In spite of the compensatory adjustments in growth between individual seminal roots of barley, the growth of barley shoots, as in wheat, was reduced when part of the root system was in compacted soil.