We review the role of anisotropic stress in controlling the growth anisotropy of stems. Instead of stress, growth anisotropy is usually considered in terms of compliance. Anisotropic compliance is typical of cell walls, because they contain aligned cellulose microfibrils, and it appears to be sufficient to explain the growth anisotropy of an isolated cell. Nevertheless, a role for anisotropic stress in the growth of stems is indicated by certain growth responses that appear too rapid to be accounted for by changes in cell-wall compliance and because the outer epidermal wall of most growing stems has microfibrils aligned axially, an arrangement that would favour radial expansion based on cell-wall compliance alone. Efforts to quantify stress anisotropy in the stem have found that it is predominantly axial, and large enough in principle to explain the elongation of the epidermis, despite its axial microfibrils. That the epidermis experiences a stress deriving from the inner tissue, the so-called 'tissue stress', has been widely recognized; however, the origin of the dominant axial direction remains obscure. Based on geometry, an isolated cylindrical cell should have an intramural stress anisotropy favouring the transverse direction. Explanations for tissue stress have invoked differential elastic moduli, differential plastic deformation (so-called differential growth), and a phenomenon analogous to the maturation stress generated by secondary cell walls. None of these explanations has been validated. We suggest that understanding the role of stress anisotropy in plant growth requires a deeper understanding of the nature of stress in hierarchical, organic structures.