The age-specific trend of cancer incidence rates, but not its magnitude, is well described employing the multistage theory of carcinogenesis by Armitage and Doll in combination with the senescence model of Pompei and Wilson. We derived empirical parameters of the multistage-senescence model from U.S. Surveillance, Epidemiology, and End Results (SEER) incidence data from 2000–2003 and 2010–2013 for The Cancer Genome Atlas (TCGA) cancer types. Under the assumption of a constant tumor-specific transition rate between stages, there is an extremely strong linear relationship (P < 0.0001) between the number of stages and the stage transition rate. The senescence tumor suppression factor for 20 non-reproductive cancers is remarkably consistent (0.0099±0.0005); however, five female reproductive cancers have significantly higher tumor suppression. The peak incidence rate for non-reproductive cancers occurs at a younger age for cancers with fewer stages and their carcinogenic stages are of longer duration. Driver gene mutations are shown to contribute on average only about a third of the carcinogenic stages of different tumor types. A tumor’s accumulated incidence, calculated using a two-variable (age, stage) model, is strongly associated with intrinsic cancer risk. During both early adulthood and senescence, the pace of tumor suppression appears to be synchronized across most cancer types, suggesting the presence of overlapping evolutionary processes.