Electropermeabilization is obtained when the membrane potential difference reaches a critical threshold. This is performed by submitting cells to an external electric field pulse. The field modulates the endogenous potential difference in a cell-size-dependent way. Computer simulations predict that large cells would be specifically permeabilized in a mixture with smaller cells. This was examined on a mixture of Chinese hamster ovary (CHO) cells and erythrocytes. CHO cells were permeabilized to Trypan blue without any occurrence of haemolysis. A similar 'size' specificity was observed on blood samples. This agreement between prediction and experimental observation indicates that induction of electropermeabilization is mainly under the control of the size of the target cell. Its physiology plays only a minor role, if any. Treating blood with 10 square wave pulses lasting 100 microseconds of an intensity of 1.6 kV/cm induced the permeabilization of 70% of the leucocytes (polymorphs and monocytes) but did not affect erythrocytes. No washing of the sample was needed in a procedure in which cells were pulsed in the plasma. A flow electropulsing process allows the treatment of large blood volumes under conditions where cells are kept viable. These results show that electropermeabilization could be used as an effective way to obtain immunocompatible drug vehicles.