The study focuses on the Arab—Israeli conflict as reflected in 26 international crises over the 1947—2000 period. It draws upon data from the ICB dataset. While the subject matter and empirical case study determine a realist orientation, the study is modulated by an appreciation for the worldwide phenomenon of ethnic conflict. The main concern is the distinction between endurance and magnitude of a conflict that has both interstate and ethnic dimensions. In contrast to existing research in international crisis in general and in the Arab—Israeli conflict in particular, the article asserts that in protracted ethnic conflicts, endurance and magnitude do not always concur. Endurance, measured by the frequency of crises, encompasses the temporal aspect of the conflict. In this study, magnitude is measured via three indicators which take place throughout an international crisis: gravity of threat, violence, and outcomes. Using theoretical postulates, it is suggested and proved that (1) ethnic—state crises will manifest themselves more than interstate ones in conflict endurance; (2) interstate crises will reveal higher magnitude than ethnic—state ones; and (3) ethnic—state crises will follow the patterns of interstate crises but with a time-lag. Hence, it is the ethnic dimension of the conflict that prohibits the dispute from conclusion. As far as magnitude over time is concerned, some decline has taken place on the interstate level of the Arab—Israeli conflict. Finally, magnitude changes in the interstate domain preceded those in the ethnic—state one. Moreover, modifications in the interstate donsain had an impact on the ethnic—state dimension present since the first Arab—Israeli crisis, and predominant since 1974.