Extreme flooding foretold by stream network organization and flow regime

River floods are the commonest natural disaster worldwide causing substantial monetary losses and death tolls yearly. Despite enormous efforts, gauging the peril of extreme floods is an outstanding challenge for varied economic sectors and the society at large. Appraisal of the flood hazard is especially difficult when the magnitude of the rarer floods strongly increases. Limited data records in fact hinder prediction of these instances and the identification of flood divides marking the rise of progressively larger floods. Here we demonstrate that the spatial organization of stream networks and the river flow regime control the appearance of flood divides and extreme floods. Unlike their ubiquitous attribution to anomalous hydrometeorological conditions, we show for the first time that the propensity to generate extreme floods is well predicted by intrinsic properties of river basins. Most importantly it can be assessed prior to the occurrence of catastrophes through measurable metrics of these properties derived from commonly available discharge data, with far-reaching implications for flood hazard assessment and preparedness.