From numerous epidemiological studies it is evidenced that risk after exposure to a carcinogen varies with time and there has been an increasing discussion about the temporal variation in case of smoking, ionizing radiation and various chemical carcinogens. The results of several independent epidemiologic studies of occupational cohorts with benzene exposure and the development of leukemia will be presented in order to find common aspects. In this context the data of 537 confirmed cases of leukemia as an occupational disease in Germany during the time period 1978–2007 will be presented. It is concluded that the epidemiologic findings are consistent and demonstrate a smaller or even absent risk of leukemia 10–15 years after exposure to benzene has been stopped. Temporal changes in relative risk highlights the importance of examining the relationship between follow-up time and risk estimates as part of the risk assessment process.