Various theories have been given for the non‐cubic anisotropy arising from preferential site ordering of rare earth ions. It is stressed that all are symmetry theories, making no essential reference to mechanisms. Gyorgy et al1have given a theory based on macroscopic crystal symmetry, and a more detailed version based on site symmetry. Rosencwaig et al2and H. Callen3have focussed on the symmetry of each successive shell of neighbors, to isolate the dominant interactions responsible for the anisotropy. Possible mechanisms are reviewed and their inferences for temperature dependence is described. Concentration dependence of the effect is also reviewed.4Aline Akselrad5has recently observed growth‐induced non‐cubic anisotropy in garnets with a single rare earth, precluding preferential ordering in dodecahedral sites, and in garnets with purely diamagnetic dodecahedral ions, precluding anisotropic exchange interactions. In these cases Akselrad and Callen5have shown that the effect may be a single‐ion anisotropy of tetrahedral iron ions, whose symmetry is lowered by preferential site ordering among the neighboring tetrahedral sites. Dodecahedral ordering has the same effect on tetrahedral iron, and it almost certainly contributes a somewhat smaller but significant anisotropy in mixed rare earth garnets. Again these alternate models of growth‐induced anisotropy are discussed in detail, stressing physical mechanisms.