In the cold winter of 1966 Aleksay Olovnikov, a theoretical biologist at the Academy of Sciences in Moscow, was waiting in the subway station where he was hit by the idea that the ends of linear chromosomes can't be replicated fully during each round of replication. In a theoretical paper (Olovnikov, 1971) he proposed that in somatic cells the ends of the chromosomes are not fully replicated during DNA synthesis, resulting in the shortening of linear DNA molecules with each cell division, and that this may be the cause of cell cycle arrest in senescent cells. Almost two decades after this proposal, Calvin Harley and co-workers found that telomeres, the physical ends of human chromosomes, shorten as a function of age in human cells in vitro and in vivo. The telomere hypothesis proposes that critically short telomeres may act as a mitotic clock to signal the cell cycle arrest at senescence (Harley, 1991). Here, we extend the telomere hypothesis and propose a model that incorporates recent advances in tumor suppressors and cell cycle control with several areas of cell aging. We propose that telomere shortening per se is not the direct signal for cell cycle arrest. It is the consequence of telomere loss, which may lead to generation of ds or ss DNA breaks. These breaks activate a p53 dependent or independent DNA-damage pathway that leads to the induction of a family of inhibitors of cyclin dependent kinases (including p21 and p16) and the eventual G1 block of senescence. In agreement with this hypothesis, we demonstrate that the level of p53 protein increases in near senescent cultures of MDFs. This increase may be responsible for induction of p21 (Noda, 1993) and IGF-Bp3 (Goldstein, 1991).