Cumulative damage models are a macroscopic approach to describe the epidemiologic aspects of carcinogenesis. This approach is based on the conjectures that cancer epidemiology (a) may derive only limited insight into specific biologic processes, and (b) may need no detailed biologic assumptions to obtain sufficient understanding of the relation between environmental exposures and host responses in human populations. The cumulative damage model leads intrinsically to the theoretical interpretation of the epidemiologic aspects of carcinogenesis as a kind of wear-out process. In the present paper, it has been applied to the data of the British physicians' study and is contrasted with some other convenient models. The comparisons show that all considered models fit the data reasonably well. The conclusion is essentially theoretical: because the present model fits the data sufficiently closely, but is based on rather parsimonious biologic assumptions, the specific referral of epidemiologic observations to cellular events, as done by other approaches, might be an overinter-pretation of the data.