Too often, flood hazard diagnostic and assessment of the effect of flood reduction strategies relied on one “design flood”, either historical or probabilistic. The 2007 European Flood Directive should help generalizing better practices, by demanding 3 probabilistic flood hazard maps in each Area of Potentially Significant Flood Risk, and by recommending economic analyses to define flood reduction strategies. However, considering 3 return periods is only a step towards real regime-wise studies. Moreover, the widely used reach-wise approaches do not take correctly into account the effect of the hydraulic infrastructures at catchment scale. Both scientific and operational communities investigate alternative methods retaining the inter- and intra-event variability, and taking into account the effect of hydraulic structures. Here, we present a chain of tools for continuous simulation. Stochastic distributed rainfall fields are fed into a rainfall-runoff model, which in turn feeds a hydraulic model. Nowadays, computing tools can cope with very long flood time-series, from which flood quantiles can be derived at any point. In its present state, our prototype can highlight the drawbacks of reach-wise approaches, and help advocate the use of this approach, for urban floods or for catchment-scale management. We shall discuss the perspectives to better take into account extreme events, and to work at larger scales, where several weather types exist on different parts of the catchment. We shall also discuss the interest of estimating expected damages from the flood hazard time series, using damage curves. From the damage time-series, Expected Annual Damages (EAD) can be derived. In theory, EAD are ideal indicators of flood risk and, by difference of scenarios, can help assess the efficiency of flood risk reduction measures but also of a whole strategy. In practise, it is important to know their drawbacks to make the best use of them, among other indicators.