The effectiveness of the adaptive immune system relies upon extensive proliferation of an initially small number of antigen-specific T cells. At the end of a successful response, the majority die by apoptosis and a small minority joins the memory cell pool. Upon re-challenge with antigen, these memory cells must again undergo clonal expansion in order to mediate an effective response. Thus, T cells are subjected to marked proliferative stress which may result in clonal exhaustion due to replicative senescence. In other systems made up of rapidly proliferating cells (e.g. in the gut) individual clones are identical and are replaced at the end of their lifespan by differentiation from a stem cell reservoir. However, because of the unique clonal distribution of antigen receptors on T cells, mere replacement with other T cells is not sufficient to maintain the integrity of the system. Moreover, the very source of new T cells decreases with age (due to thymic involution). Therefore, the adaptive immune system may be uniquely susceptible to the deleterious effects of replicative senescence. Particularly in humans, in vivo studies of the behaviour of individual T-cell clones in the body is difficult. However, T-cell longevity, measured as proliferative capacity in terms of population doublings, can be usefully modelled at the clonal level in vitro. This paper discusses the surprisingly little that is known about the average longevity, variation between clones, and the maximal longevity of human T cells under clonal culture conditions in vitro. From our own studies, we show that average lifespan of human T cells is as little as 17 PD; however, established clones reach 35 PD on average, with maximum longevity generally in the region of 60-80 PD, regardless of the source of the cloned cells. Expression of surface molecules in general did not differ strikingly between young and old donors, but the frequency of clones secreting IL-10, and the amount secreted per clone was higher in the elderly than in the young. Conversely, the frequency of clones secreting IL-6 and the amount secreted per clone was higher in the young.