The objective of the paper is to provide a systematic consideration and generalization of properties and features of the FDTD method in the context of its use in solving microwave power problems. This is aimed at filling the gap between the general theory of the FDTD method and the specific practice of its applications by microwave power engineers. The paper starts with a comparison of FDTD to other methods like FEM, from the perspective of microwave power simulations. It then discusses FDTD-specific models of lossy and dispersive media, conformal boundaries, field singularities, and modal excitation as well as error bounds due to numerical dispersion. Theoretical overview is illustrated with examples. References are provided to the literature where more details and application notes can be found.