Substantial numbers of the world’s chronically poor and malnourished population live in an increasingly volatile world. The dangerous nexus of climate change, rapid population growth, conflict, and food price volatility already appears to have pushed several poor regions into states of permanent crisis, even as the rest of the world has seen unprecedented progress against poverty. This disturbing state of affairs, along with our expanded knowledge of the intimate interactions between short-term shocks and long-run development, has catalyzed widespread interest in resilience building, and in what such a framework implies for understanding the causes and consequences of acute vulnerability to natural and man-made disasters. In this paper we ask what this paradigm implies for the measurement and analysis of resilience. Resilience is fundamentally about complex dynamics. Slower-moving ecological, economic, demographic, and social stressors create vulnerability to short-run shocks, which in turn can have long-term consequences by reinforcing preexisting vulnerabilities. In our view, this basic conception of resilience has fundamental measurement implications. First, resilience can be measured and understood only through higher-frequency surveys that capture the causes and consequences of time-varying stressors and shocks, including seasonal shocks. Second, resilience can be understood only through surveys that capture the multidimensional complexity of stressors, shocks, and feedback loops, including the complex interactions between economic, social, and ecological forces. Third, the underlying stressors that create vulnerability, and the resilience-building interventions that reduce vulnerability, can be gauged and evaluated only over the longer term. This conceptualization of resilience motivates us to go a step further than existing research on resilience and on food and nutrition security measurement, by outlining a far more expansive strategy for improving and scaling up the monitoring, measurement, and analysis of the world’s most vulnerable populations. We propose the development of a multicountry system of high-frequency, long-term sentinel sites in the world’s most vulnerable regions. If implemented along the lines we conceive, this system could be a high-return investment for resilience-building efforts, since it would serve multiple purposes. This system offers the only rigorous means of monitoring vulnerability and resilience in the world’s most volatile regions. This system would bolster existing early-warning systems by complementing them with household-level indicators. This system would improve the targeting of emergency resources. This system would be instrumental for diagnosing the underlying sources of vulnerability, for identifying key thresholds of resilience, and for designing appropriate resilience-building strategies. And this system would provide a rigorous foundation for large-scale evaluations of resilience-building activities. While there are strong justifications for such a system, the devil is necessarily in the details, and much of this paper is concerned with those details. Largely to learn from existing experience, we first review existing measurement strategies that are similar in purpose or design to the sentinel system outlined above. When implemented, long-term, high-frequency measurement systems have often yielded great benefits but been hampered by cost, lack of institutional coordination, and insufficient dissemination and usage of data. The need to keep costs down and benefits widespread therefore motivates us to consider which countries in the world have the highest priority for the development of sentinel sites, based on indicators such as child nutrition and health outcomes, exposure to disasters, and past emergency assistance levels from the international community. We then turn to crucial issues of data collection design by outlining a hybrid sampling and survey design that will help achieve the various objectives outlined above while keeping costs down. We also argue that the proliferation of mobile phones and other information and communications technologies offers substantial scope for a cost-effective system of this kind, far more so than would have been available in the past. Finally, we consider who should lead and contribute to this ambitious effort. Since the principal advantage of this approach is that it can yield benefits for a wide range of institutions and purposes (relief and development, operations and research, social and biophysical sciences), and since the costs of a long-term commitment to these sentinel surveys would be large indeed for any single agency, we propose the need for a relatively broad consortium of international donors. This consortium should first focus on establishing partnerships with national governments and then commit to long-term resilience monitoring as well as domestic capacity building. With this essential commitment in place, this consortium would then need to secure implementing partners with a permanent presence on the ground, as well as the technical expertise of international organizations of various sorts. Ultimately, we argue, it is only this kind of long-term, cooperative commitment that will provide a scientific evidence base for diagnosing and resolving the world’s worst problems of hunger, poverty, and malnutrition. Only this kind of sentinel system can generate the data and evidence needed to inform actions to build resilience and to help the global community eliminate extreme poverty in the generation ahead. The status quo is simply not enough.