Clonal hemopoiesis driven by leukemia-associated gene mutations can occur without evidence of a blood disorder. To investigate this phenomenon, we interrogated 15 mutation hot spots in blood DNA from 4,219 individuals using ultra-deep sequencing. Using only the hot spots studied, we identified clonal hemopoiesis in 0.8% of individuals under 60, rising to 19.5% of those ≥90 years, thus predicting that clonal hemopoiesis is much more prevalent than previously realized. DNMT3A-R882 mutations were most common and, although their prevalence increased with age, were found in individuals as young as 25 years. By contrast, mutations affecting spliceosome genes SF3B1 and SRSF2, closely associated with the myelodysplastic syndromes, were identified only in those aged >70 years, with several individuals harboring more than one such mutation. This indicates that spliceosome gene mutations drive clonal expansion under selection pressures particular to the aging hemopoietic system and explains the high incidence of clonal disorders associated with these mutations in advanced old age. Display omitted •Clonal hemopoiesis is an almost inevitable consequence of aging in humans•Spliceosome gene mutations drove clonal hemopoiesis only in persons aged ≥70 years•NPM1 mutations behave as gatekeepers for leukemogenesis McKerrell et al. employ ultra-deep sequencing to show that age-related clonal hemopoiesis is much more common than previously realized. They find that clonal hemopoiesis, driven by mutations in spliceosome genes SF3B1 and SRSF2, was noted exclusively in individuals aged 70 years or older and that NPM1 mutations are not seen in association with this phenomenon, endorsing their close association with leukemogenesis.