Age‐adjusted cancer mortality data (1964) were examined for evidence of independence. Indications were found that substitution of one cancer for another is a common occurrence. The data were interpreted as supportive of the view that natural selection for resistance to cancer in general has occurred with many resistance‐related genes common to prevention of cancer of several sites. Consequently, the comparison of age‐adjusted incidence or mortality rates of cancer of single sites alone does not provide a satisfactory estimate of the magnitude of an “environmentally induced” cancer risk. It is necessary to examine all causes of death to find indications of a real life‐shorten