The overall cancer incidence rate declines at very old age. Possible causes of this decline include the effects of cross-sectional data that transform cohort dynamics into age patterns, population heterogeneity that selects individuals susceptible to cancer, a decline in some carcinogenic exposure in older individuals, underdiagnosis, and the effects of individual aging that slow down major physiological processes in an organism. Here several mathematical models contributing to the explanation of this phenomenon are discussed, with extension of the Strehler and Mildvan model of aging and mortality to the analysis of data on cancer incidence at old age (data source: International Agency for Research on Cancer). The model can help explain the observed time trends and age patterns of cancer incidence rates.