At the heart of the traditional approach to strategy in the climate change dilemma liesthe assumption that the global community, by applying a set of powerful analyticaltools, can predict the future of climate change accurately enough to choose a clearstrategic direction for it. We claim that this approach might involve underestimatinguncertainty in order to lay out a vision of future events sufficiently precise to becaptured in a discounted cost flow analysis in integrated assessment models. However,since the future of climate change is truly uncertain, this approach might at best bemarginally helpful and at worst downright dangerous: underestimating uncertainty canlead to strategies that do not defend the world against unexpected and sometimes evencatastrophic threats. Another danger lies on the other extreme: if the global communitycan not find a strategy that works under traditional analysis or if uncertainties are toolarge that clear messages are absent, they may abandon the analytical rigor of theirplanning process altogether and base their decisions on good instinct and consensus ofsome future process that is easy to agree upon.In this paper, we try to outline a system to derive strategic decisions under uncertaintyfor the climate change dilemma. What follows is a framework for determining the levelof uncertainty surrounding strategic decisions and for tailoring strategy to thatuncertainty.Our core argument is that a robust strategy towards climate change involves thebuilding of a technological portfolio of mitigation and adaptation measures that includessufficient opposite technological positions to the underlying baseline emission scenariosgiven the uncertainties of the entire physical and socioeconomic system in place. In thecase of mitigation, opposite technological positions with the highest leverage areparticular types of sinks. A robust climate risk management portfolio can only workwhen the opposite technological positions are readily available when needed andtherefore they have to be prepared in advance. It is precisely the flexibility of thesetechnological options which has to be quantified under the perspective of the uncertainnature of the underlying system and compared to the cost of creating these options,rather than comparing their cost with expected losses in a net present value typeanalysis. We conclude that climate policy ― especially under the consideration of theprecautionary principle ― would look much different if uncertainties would be takenexplicitly into account.