The theory and practice of investigations on cancer clustering were reviewed. The main topics reviewed were the following: history of the theory of disease cluster analysis, the practical experience acquired by the National Institute of Occupational Safety and Health (NIOSH), theory versus practice, and the perceptions of workers and employees. Between 1978 and 1984, NIOSH carried out a total of 61 investigations of apparent cancer clusters, most of which consisted of 5 cases or less. None of these clusters showed evidence of occupational etiology. Although known or suspected carcinogens were identified in 43 percent of the investigations conducted during this period, no individual substance prevailed and no relationship could be established between the extent of the exposure and the risk of disease among the members of a given cancer cluster. The most common problems encountered during these investigations were the small number of cases in occupational settings, the lack of complete personnel records and, at times, because of the condition of the records or the nature of the request, the identification of the population at risk. The authors point out the fact that the traditional epidemiological techniques frequently cannot be applied to the study of apparent cancer clusters occurring in occupational settings.