Large-scale war is a world-system phenomenon of the rivalry phase. Such conflicts have once again become a concern, and nuclear weapons make these prospects especially dangerous. This is particularly problematic since several world-systems perspectives suggest the chances for war will be greatest in the period from 2030 to 2050. I review the logic of rivalry, the reasons for the endurance of nuclear weapons, old and new nuclear strategies, and the processes that may pose the greatest existential dangers. Chase-Dunn and Podobnik (1995) identified processes that militate both in favor of and in opposition to nuclear war, and I pay particular attention to the way in which world-systems developments that increase the likelihood of major war have persisted, while those that retard the chances for major war have diminished. These dangers suggest that it may be time to turn some of our attention to the dynamics of systemic war and nuclear weapons.