Abstract Flooding can result in the temporary or permanent displacement of affected populations from their homes. This displacement is not only an important factor determining health and social impacts, but it also has bearing on the monitoring of such impacts. We present descriptive analyses of data from the 2007 floods of Kingston‐upon‐Hull, UK, which quantify the frequency, timing and duration of household displacements. Very high proportions (59%) of flooded households were displaced, though many displacements did not begin until several months after flooding. Twelve per cent of displaced households remained out of their homes for longer than a year, a further 5% for more than 2 years. These figures suggest that linkages between a flood incident, flood impacts and then recovery might not always follow a neat linear progression, which has important implications for epidemiological research design as well as flood risk management plans. This is the first reported investigation on population displacement after the flood event in the UK.