The observation that victimisation raises the probability of subsequent victimisation is well established. That repeat crimes tend to happen quickly is important in the preventive utility of the phenomenon. However, the time decay curve of repeat victimisation is conventionally studied using police recorded crime data, making it susceptible to alternative explanations by confounding actual victimisation with rates of victim report to the police. The present paper reports an attempt to use victimisation survey data to demonstrate decay curves. The analyses show that households experiencing a high frequency of repeat victimisation experience repeats much sooner than households where repeats occur fewer times. Subsequent repeats take place ever more quickly after the preceding one. Quick victimisation is perceived to be increasingly serious the sooner it occurs.