"Immunosenescence" is a loosely descriptive designation for age-associated alterations to most measures of immunity, which can be seen in all mammals examined in any detail. Both innate and adaptive immunity are affected in a manner assumed to be deleterious, but often the clinical consequences of the assessed changes are unclear or not even investigated. The mechanisms accounting for these changes, and biomarkers of immunosenescence, are currently the subject of intensive research. Cross-sectional studies have established hallmark age-associated differences between adaptive immune factors in young and old people, particularly a lower number and percentage of naïve T cells, especially CD8+ T cells, and accumulations of late-differentiated CD8+ T cells. The latter but not the former is strongly affected by infection with the persistent ß-herpesvirus HHV5 (cytomegalovirus, CMV). Only limited longitudinal studies have so far investigated whether these differences actually reflect age-associated changes at the individual level. The Swedish OCTO/NONA-Immune studies identified a set of immune parameters including infection with CMV which predicted survival in people over 85 at baseline. Moreover, the Leiden 85+ study showed that T cell-mediated pro-inflammatory specific for CMV antigens was enriched in very old survivors, suggesting the overarching necessity of maintaining effective immunosurveillance of this virus. Here, the disparate impact of CMV on "immunosenescence" and survival in human populations under different condition is reviewed.