A theoretical examination of explosive crater scaling rules based upon dimensional analysis is presented. The distinction between the scaling of similar experiments as opposed to the more common practice of scaling nonsimilar experiments is emphasized. Three different sets of dimensionless &pgr; groups commonly used in the literature (denoted here as the mass set, the energy set, and the gravity set) are compared, suggesting alternate ways to preserve similarity among experiments. Specific scaling rules using an assumed power law functional relationship among &pgr; groups are derived for nonsimilar experiments. These scaling rules result from assumptions that cratering is independent of certain variables. The role of explosive energy density is examined and shown to be an essential variable which must be included in the analysis. None of the derived results lead exclusively to the requirement of cube root or quarter root scaling except under special circumstances, which are shown to be common to all three formulations. In the usual series of experiments conducted in the same soil type with an identical type of explosive the scaling rules are shown to be bounded by cube root and quarter root rules. Other scaling exponent restrictions are presented.