Worst-case or high-end subduction-related earthquakes and tsunamis of 2004 and 2011 are painfully fresh in our memories. High-end subductionrelated volcanic eruptions have not occurred in recent memory, so we review historical and geologic evidence about such eruptions that will surely recur within coming centuries. Specifically, we focus on Volcanic Explosivity Index (VEI) 7 eruptions, which occur 1-2 times per thousand years. A variety of environmental changes followed the VEI 7 eruption of Rinjani (Samalas), Indonesia, in CE 1257 and several more eruptions of VEI 6 or 7 that occurred in the succeeding few centuries. The Rinjani eruption and its impacts are relatively well documented, modeled, and, for the purposes of attribution, uncomplicated by antecedent eruptions. It seems likely that the Rinjani eruption triggered the onset of the Little Ice Age, and subsequent large eruptions sustained it. Although climatic effects of eruptions like Pinatubo (Philippines) and even Tambora (Indonesia) lasted only a few years, it seems that coupling of oceans, sea ice, and atmosphere after larger eruptions can force decade-to century-long cooling, decreased precipitation, and major impacts on crops. The next VEI 7 will affect a world very different from that of CE 1257. Today, populations within 100 km of candidate volcanoes range from fewer than 1000 people in remote areas to between 20 and 30 million people near several candi dates in Indonesia and the Philippines. If a VEI 7 eruption occurs, those populations will be at dire risk, and eruptions in some locations could destabilize financial centers, national economies, and even peace between nations. Distal effects on air travel, the global positioning system, and climate will be felt by a high-technology, globally interdependent world. We suggest and apply criteria to identify candidates for future VEI 7 eruptions, and discuss likely challenges for short-range forecasting of such events. Preparation for such low-probability but high-consequence events is difficult to imagine, yet some modest early measures can be considered. Volcanologists should refine geologic histories and ensure at least baseline monitoring of candidate volcanoes, and consider how they will judge the likelihood that an impending eruption will be of VEI 7. Forward-thinking governments and industries would be wise to consider how a proximal or distal VEI 7 eruption will affect their jurisdictions, and what responses make the most economic and sociopolitical sense. SIZES OF EXPLOSIVE VOLCANIC ERUPTIONS An imperfect but handy measure of explosive magnitude is the VEI (Newhall and Self, 1982), a broadly logarithmic scale similar to Richter magnitude for earthquakes and to a more precise but also more data-demanding magni-GEOSPHERE