Instead of assuming that all actors are equally likely to clash, and that they do so independently of previous clashes, rivalry analysis can focus on the small number of feuding dyads that cause much of the trouble in the international system. But the value added of this approach will hinge in part on how rivalries are identified. Rivalry dyads are usually identified by satisfying thresholds in the frequency of militarized disputes occurring within some prespecified interval of time. But this approach implies a number of analytical problems including the possibility that rivalry analyses are simply being restricted to a device for distinguishing between states that engage in frequent and infrequent conflict. An alternative approach defines rivalry as a perceptual categorizing process in which actors identify which states are sufficiently threatening competitors to qualify as enemies. A systematic approach to identifying these strategic rivalries is elaborated. The outcome, 174 rivalries in existence between 1816 and 1999 are named and compared to the rivalry identification lists produced by three dispute density approaches. The point of the comparison is not necessarily to assert the superiority of one approach over others as it is to highlight the very real costs and benefits associated with different operational assumptions. The question must also be raised whether all approaches are equally focused on what we customarily mean by rivalries. Moreover, in the absence of a consensus on basic concepts and measures, rivalry findings will be anything but additive even if the subfield continues to be monopolized by largely divergent dispute density approaches.