Shear deformation of partially molten rock in laboratory experiments causes the emergence of melt-enriched sheets (bands in cross-section) that are aligned at about 15-20 • to the shear plane. Deformation and deviatoric stress also cause the coherent alignment of pores at the grain scale. This leads to a melt-preferred orientation that may, in turn, give rise to an anisotropic permeability. Here we develop a simple, general model of anisotropic permeability in partially molten rocks. We use linearised analysis and nonlinear numerical solutions to investigate its behaviour under simple-shear deformation. In particular, we consider implications of the model for the emergence and angle of melt-rich bands. Anisotropic permeability affects the angle of bands and, in a certain parameter regime, it can give rise to low angles consistent with experiments. However, the conditions required for this regime have a narrow range and seem unlikely to be entirely met by experiments. Anisotropic permeability may nonetheless affect melt transport and the behaviour of partially molten rocks in Earth's mantle. arXiv:1505.00559v2 [physics.geo-ph] 5 Sep 2015 whether the low angles could arise as a consequence of a permeability that is directionally dependent (i.e. anisotropic).Anisotropic permeability could be a consequence of another empirically known feature of partially molten rocks subjected to deviatoric stress: the microstructural alignment of interconnected pockets of melt between solid grains. This is called melt-preferred orientation (MPO) and has been observed in many laboratory studies [e.g. Bussod and Christie, 1991, Daines and Kohlstedt, 1997, Takei, 2010. The alignment may be attributed to the instantaneous state of deviatoric stress [Daines and Kohlstedt, 1997, Takei andHoltzman, 2009a], or to the combined effects of finite strain, lattice-preferred orientation, and anisotropic surface energy of olivine grains [Bussod and Christie, 1991, Daines and Kohlstedt, 1997, Jung and Waff, 1998; it is likely some combination of the two. Since the Darcian permeability of partially molten rocks arises from the shape and interconnectedness of melt pockets at the grain scale [e.g. Bear, 1972, Scheidegger, 1974, it is reasonable to assume that the anisotropic alignment of pores between grains leads to anisotropy in permeability. Daines and Kohlstedt [1997] estimated this anisotropy as a function of differential stress and found that permeability in the direction parallel to the maximum compressive stress σ 1 was enhanced by a factor of up to 15 over that parallel to the direction of maximum tensile stress. This is consistent with a theoretical model for anisotropy of permeability due to MPO by Hier-Majumder [2011].Since both melt-banding at the macroscopic scale and melt-preferred orientation at the microscopic scale emerge under the same physical conditions, it is logical to ask whether their dynamics are linked. In particular, the question we address here is whether the low angle of high-porosity bands observed in experiments [see s...