Magnetic polymer-coated particles linked to antibodies are considered to be an efficient rosetting matrix for immunoselection. We have shown that a 20:1 bead:target cell ratio and a 90 min incubation period are the optimal conditions for specific binding of monoclonal antibody-labelled cells to goat anti-mouse IgG-coated beads. Higher ratios or longer incubation periods resulted in considerable non-specific binding. Characterisation of the optimal conditions for specific depletion of lymphocyte subpopulations showed that (a) a range of bead:target cell ratios and incubation periods can be used, with resulting high efficiency and specificity; (b) multiple monoclonal antibodies can be used simultaneously for the depletion of diverse lymphocyte subpopulations; (c) non-specific bead-to-cell binding does not affect the specificity and efficiency of magnetic depletion; (d) specific binding of one bead only was adequate for effective magnetic separation. These findings define the most economical, specific and efficient conditions of use of beads for negative immunoselection but preclude the use of beads as an analytical rosetting medium.