Selected aspects of magnetic surface anisotropy are reviewed. The emphasis is on methods for deducing reliable surface anisotropy values from experiments such as ferromagnetic resonance at microwave frequencies and Brillouin scattering at optical frequencies. The methods used are the "general exchange boundary condition method" and the "effective volume anisotropy method". The essence of the former is the supplementing of the equation of motion of the magnetization with the general exchange boundary condition whereas the latter consists of using the "stratagem" of effective volume anisotropy. We find that use of the general exchange boundary condition method is not only preferable in principle but often actually necessary to prevent the prediction of wrong surface anisotropy values and to permit the prediction of some observable Brillouin shifts.