Since the first report of a possible cluster of childhood leukaemia in 1963, many studies have shown a tendency to clustering in space and/or in time of this disease. This paper summarizes the studies about clusters of childhood leukaemia in the international context and in Italy, and discusses the aetiological hypotheses particularly in reference to exposures in the area of residence. Thirty-three cluster studies have been identified through a literature search performed using the PubMed bibliographic database, websites of main national and international health authorities, and relevant reviews published on this topic. Most of leukaemia cluster studies were conducted following an alarm in the nearby areas of a point source of toxic substances as nuclear power plants, chemical-industrial sites, military bases, broadcasting antennas. As alternative explanation, in several clusters an infectious aetiology was hypothesized (population mixing or delayed infection hypothesis). The uncertainty associated with cluster investigations underlines the importance of applying rigorous methods and transparent decision making, involving parents of children with cancer and the local community, and taking into account all potential aetiological factors, especially those for which a biological plausibility exists.