The human immune system evolved to defend the organism against pathogens, but is clearly less well able to do so in the elderly, resulting in greater morbidity and mortality due to infectious disease in old people, and higher healthcare costs. Many age-associated immune alterations have been reported over the years, of which probably the changes in T cell immunity, often manifested dramatically as large clonal expansions of cells of limited antigen specificity together with a marked shrinkage of the T cell antigen receptor repertoire, are the most notable. It has recently emerged that the common herpesvirus, cytomegalovirus (CMV), which establishes persistent, life-long infection, usually asymptomatically, may well be the driving force behind clonal expansions and altered phenotypes and functions of CD8 cells seen in most old people. In those few who are not CMV-infected, another even more common herpesvirus, the Epstein-Barr virus, appears to have the same effect. These virus-driven changes are less marked in "successfully aged" centenarians, but most marked in people whom longitudinal studies have shown to be at higher risk of death, that is, those possessing an "immune risk profile" (IRP) characterized by an inverted CD4:8 ratio (caused by the accumulation primarily of CD8(+) CD28(-) cells). These findings support the hypothesis that persistent herpesviruses, especially CMV, act as chronic antigenic stressors and play a major causative role in immunosenescence and associated mortality.