The response of roots to mechanical impedance has been addressed in the literature largely from the physical point of view. The properties of soils which cause them to become impenetrable by roots have been analysed in detail, with particular reference to soil texture. Factors such as high soil cohesion (in clay soils) and high angle of internal friction (in sandy soils) contribute to soil strength. However, root growth often involves radial deformation of the soil near the growing apex, requiring a consideration of soil compression as well. While soils of all textures can impede root growth, those with high clay content are thought to be most inhibitory. Predictions of soil strength can also be obtained from penetrometer probes with different diameters and tip shapes. A precise physical analogue of root growth is not possible but probes which penetrate soil by deformation around the tip give surprisingly good estimates of relative soil strength. The capacity of roots to minimize friction with the soil and expand radially is thought to account for the lower absolute resistance perceived by roots than by penetrometer probes. Roots oppose strong soil by forces of osmotic origin acting on both the soil and the expanding cell walls. The response of roots is, however, poorly understood. Cortical cells tend to become broader and shorter, causing the root axis to thicken. Root volumes and osmotic pressures changes as a result. The role of ethylene as a mediator of structural changes is in question. Root (and shoot) carbohydrate metabolism is also change by impedance in a way that produces a favourable balance of biomass above and below ground and prevents carbohydrate deprivation to growing tissues. However, the co-ordination of changes in anatomy and metabolism remains a mystery. The scope for selection of plants tolerant to mechanical impedance is discussed and there are reasons for optimism if new screening criteria are adopted.