The purpose of the paper will be to engage with concepts that are widely used in the discourse on water and risk; that is vulnerability, resilience, uncertainty and surprises. The paper wants to establish some linkages between these concepts both in a conceptual and in an empirical sense. It departs from outlining the central challenges associated with the enormous uncertainties connected with projecting the consequences of climate change. In a second step it introduces Wildawsky’s differentiation between the concepts of anticipation and resilience to outline two different ways of how to deal with such uncertainties (1991). While anticipation assumes to know the future; resilience expects – paradoxically – the unexpected. In this line, the concept of resilience seems more relevant in developing an adaptation strategy, which is aware of uncertainties about future developments. Yet, an empirical case-study on the 2002 Mulde flood (Saxony, Germany) will reveal that strategies of anticipation are by far more dominant then resilience-based strategies both on the side of citizens and decision-makers. Based on the previous argument the paper outlines a fundamental paradox: While a resilience-based adaptation strategy seems appropriate to adapt to unexpected developments, it is confronted with a deficit of implementation and acceptance (it is a rather unworldly approach). An anticipation-based strategy, on the other hand, is the accepted and dominant adaptation strategy and at the same time produces the condition for an increasing vulnerability (since it assumes to have valid knowledge about the future). The paper concludes with elaborating a possible strategy that is aware of the paradox that resilience is on the one hand desirable but not easy to implement and anticipation on the other hand not desirable but established practice.