The differential expression of CD45 isoforms has been suggested as a marker for stages of post-thymic T cell development, that is, CD45RA+CD45R0-T cells and CD45RA-CD45R0+ T cells are supposed to be virgin and memory cells respectively. Recently, several adhesion molecules have been shown to be up-regulated on the cell surface of memory T cells, and have been suggested to serve as a memory marker. In this study, we investigated the levels of LFA-1 expression on T cells in various subpopulations defined by CD45 isoform expression in donors of various ages. In CD4+ T cells, the proportion of LFA-1high cells among CD45RAhighCD45R0- T cells remained low in all age groups and did not show significant accumulation with age. CD4+CD45RA-CD45R0high T cells expressed LFA-1 at a higher level than CD4+CD45RAhighCD45R0- T cells. Thus, the currently prevailing view that CD45RA and CD45R0 can be markers for virgin and primed cells was consistent with LFA-1 expression in CD4+ T cell population. In CD8+ T cells, however, CD45RAhighCD45R0- T cells consisted of two distinct subpopulations, LFA-1low and LFA-1high cells, whereas CD45RA-CD45R0high T cells were almost exclusively LFA-1high. When CD29 expression was examined in place of LFA-1 expression, similar results were obtained; CD45RAhigh CD45R0- T cells consisted of two distinct subpopulations, CD29-to low and CD29high cells, while CD45RA-CD45R0high T cells were mostly CD29high. The proportion of LFA-1high cells in the CD8+CD45RAhigh T cell subpopulation increased significantly as a function of age (r = 0.9, p < 0.001). It ranged from 8% in a 14-year-old donor to 94% in a 79-year-old donor. Furthermore, the proportion of CD8+CD45RAhighLFA-1high cells in the CD8+ T cell population increased significantly as a function of age (r = 0.85, p < 0.001). On the other hand the proportion of LFA-1high cells in CD8+CD45RA- T cell subpopulation exceeded 90% in most donors irrespective of age. These results indicate that the CD8+CD45RAhighCD45R0- T cell subpopulation contains a considerable number of LFA-1high cells and CD29high cells, phenotypically similar to previously activated cells. Thus, in terms of LFA-1 and CD29 expressions, the simple scheme that CD45RA is a marker of virgin cells is not applicable to the CD8+ T cell population.