The interaction of rat lymphocyte subpopulations with the surfaces of poly(2-hydroxyethyl methacrylate)-graft-polyamine (HA) copolymer and poly(2-hydroxyethyl methacrylate)-graft-partially quaternized polyamine (HQA) copolymer was evaluated under a controlled shear stress by using a plane-parallel-type channel equipped with a polymer-coated bottom wail [hybrid field-flow fractionation/adsorption chromatography (FFF/AC) technique]. Estimation of the recovery profile of lymphocytes from KA and HQA channels under varying shear stress indicated that the contribution of electrostatic interaction to cellular attachment was different among lymphocyte subpopulations, having a good correlation with their apparent isoelectric points. Extended settling time caused a progressive increase in the attachment force between the lymphocyte and the copolymer surface for the HA channel, resulting in a considerable decrease in the recovery of both B and T cells from the channel, On the contrary, recovery of T cells from the HQA channel showed only a slight decrease with a settling time, allowing attainment of a T-cell population as the effluent in good yield even under the condition of prolonged settling time, The time-dependent increase in the cell/surface interaction observed for the HA channel might be due to a progressive contribution of the hydrophobic interaction caused by an increase in the hydrophobicity of the HA copolymer through the deprotonation of ethylenediamine units in polyamine grafts at the cellular interface. This assumption was supported by the fact that polyamine macro-monomer was quantitatively extracted into n-octanol from water through deprotonation at the critical pH, where the evaluation of cell/surface interaction was carried out. These results indicated that lymphocyte attachment is able to be controlled by changing the amount and nature (fixed or unfixed) of the charges in polyamine-graft copolymers.