Normal somatic cells have a finite life span  and lose telomeric DNA, present at the ends of chromosomes, each time they divide as a function of age in vivo or in culture [2-4]. In contrast, many cancer cells and cell lines established from tumours maintain their telomere length by activation of an RNA-protein complex called telomerase, an enzyme originally discovered in Tetrahymena , that synthesizes telomeric repeats [6-8]. These findings have led to the formation of the 'telomere hypothesis', which proposes that critical shortening of telomeric DNA due to the end-replication problem  is the signal for the initiation of cellular senescence [10,11]. In yeast, the EST2 gene product, the catalytic subunit of telomerase, is essential for telomere maintenance in vivo [12-14]. The recent cloning of the cDNA encoding the catalytic subunit of human telomerase (hTERT) [15,16] makes it possible to test the telomere hypothesis. In this study, we expressed hTERT in normal human diploid fibroblasts, which lack telomerase activity, to determine whether telomerase activity could be reconstituted leading to extension of replicative life span. Our results show that retroviral-mediated expression of hTERT resulted in functional telomerase activity in normal aging human cells. Moreover, reconstitution of telomerase activity in vivo led to an increase in the length of telomeric DNA and to extension of cellular life span. These findings provide direct evidence in support of the telomere hypothesis, indicating that telomere length is one factor that can determine the replicative life span of human cells.