Cultured human fibroblasts undergo senescence (a loss of replicative capacity) after a uniform, fixed number of ∼50 population doublings, commonly termed the Hayflick limit. It has been long known from clonal and other quantitative studies, however, that cells decline in replicative capacity from the time of explantation and do so in a stochastic manner, with a half-life of only ∼8 doublings. The apparent 50-cell doubling limit reflects the expansive propagation of the last surviving clone. The relevance of either figure to survival of cells in the body is questionable, given that stem cells in some renewing tissues undergo >1,000 divisions in a lifetime with no morphological sign of senescence. Oddly enough, these observations have had little if any effect on general acceptance of the Hayflick limit in its original form. The absence of telomerase in cultured human cells and the shortening of telomeres at each population doubling have suggested that telomere length acts as a mitotic clock that accounts for their limited lifespan. This concept assumed an iconic character with the report that ectopic expression of telomerase by a vector greatly extended the lifespan of human cells. That something similar might occur in vivo seemed consistent with initial reports that most human somatic tissues lack telomerase activity. More careful study, however, has revealed telomerase activity in stem cells and some dividing transit cells of many renewing tissues and even in dividing myocytes of repairing cardiac muscle. It now seems likely that telomerase is active in vivo where and when it is needed to maintain tissue integrity. Caution is recommended in applying telomerase inhibition to kill telomerase-expressing cancer cells, because it would probably damage stem cells in essential organs and even increase the likelihood of secondary cancers. The risk may be especially high in sun-exposed skin, where there are usually thousands of p53-mutant clones of keratinocytes predisposed to cancer.