Two positive selection methods were compared for the ability to capture both the bright and dim subsets of CD8 lymphocytes in mononuclear cell (MC) preparations from ten healthy individuals. The first method utilized anti-CD8-coated magnetic beads; captured cells were then recovered using a polyclonal sheep anti-mouse Fab reagent. At all bead: CD8 cell ratios tested (4:1, 8:1, 16:1), the selected cells were > 94% CD8+, and these CD8 cells were enriched for CD8bright cells (77-85%) when compared to CD8 cells in the starting MC preparation (68%). The second method utilized anti-CD8-coated culture flasks; captured cells were recovered by physical dislodgement. The recovered cells were > 90% CD8+, and these CD8 cells were modestly enriched for CD8dim cells (52%) compared to starting CD8 cells (32%). To further enrich for CD8dim cells, we used these two methods in tandem (n = 10). MC were first incubated with anti-CD8-coated magnetic beads (4:1 ratio) to obtain a CD8bright-enriched population (97% of all cells CD8+, 83% of all cells CD8bright). Uncaptured cells were incubated with anti-CD4-coated magnetic beads, and the uncaptured cells from this step were then placed in an anti-CD8-coated flask. The recovered flask-selected cell population was highly enriched for CD8dim cells (87% of all cells CD8+, 85% of all cells CD8dim). CD8 cells in the CD8bright population were 94% CD3+ and 6% CD16+, whereas those in the CD8dim population were 29% CD3+ and 66% CD16+. In proliferative studies, CD8bright cells were preferentially activated by immobilized anti-CD3, whereas CD8dim cells were preferentially activated by exogenous IL-2. In assays of natural killer activity, CD8dim cells were markedly more active than CD8bright cells. This method provides an alternative to cell sorting for obtaining enriched populations of CD8bright and CD8dim lymphocytes.