The difficulties with the geometric optics (“classical”) definition of virtual height are discussed. The more general definition of virtual height in terms of the derivative of the phase of the reflection coefficient with respect to frequency is derived. It is then shown that the former definition can be derived from the latter, if phase integral method is used. The two definitions are compared in the specific examples of linear, rectangular, Epstein, and parabolic charge distributions.it is demonstrated, by means of examples, that the relation between virtual height and frequency derivative of phase is not valid when the reflected wave contains more than one pulse. In this case the frequency derivative of phase cannot be interpreted as the time delay of any one of the pul