Application of the finite element method to the solution of physical problems is based on minimization of energy; in the present case electromagnetic energy is minimized. Representation of a volume of space by a number of finite elements and description of field or potential distribution by a finite set of unknown values make it possible to replace the energy variational equation by matrix equations. It is shown that a solution for secondary rather than total field quantities can be obtained directly. Such a procedure has several advantages.Approximations are involved in using non-infinitesimal elements and finite meshes of elements. It is usually necessary to pay more attention to mesh size than texture (element size).Examples of induced polarization anomalies over two-dimensional models illustrate effects of topography and of a highly conducting layer above bodies of polarizable material. Computed electromagnetic anomalies of two-dimensional structures, with line source excitation, include the effects of adjacent conductors and magnetic conductors set in a less conductive half-space.