We have carefully measured the magnetic anisotropy of samples cut from (112) and (110) growth surfaces of flux grown Gd 2.34 Tb .66 Fe 5 O 12 . The experimental technique included the use of spherical samples, precision x‐ray orientation, correction for the non‐collinearity of M and H, and the measurement of torque about all 〈110〉 axes. Energy curves were analyzed in terms of the most general second and fourth order terms. We find: 1. The room temperature anisotropy consists only of a cubic K1 term and a growth induced anisotropy quadratic in the direction cosines; 2. Energy surfaces of two distinctly different orientations are needed to describe the growth anisotropy under different areas of the (110) face; and, 3. Characterization of the growth induced anisotropy in terms of the two parameter model is less straightforward than has appeared to date.