We provide a novel method to discriminate the cells binding with molecules immobilized on a glass substrate by using the repulsive force of negative dielectrophoresis. The device consists of a lower substrate with interdigitated (IDA) patterns (12 μm in width and 50 μm in gap) of indium-tin oxide (ITO) and an upper ITO substrate disposed on the lower substrate. When an AC voltage is applied to a pair of microband electrodes comprising IDA to the upper ITO substrate connected to the ground, the cells are pressed onto gaps between the microband electrodes, and adhesion of the cells to the substrate is promoted (1st patterning in Fig. A). Subsequently, when an AC voltage is microband electrode “a” against the microband “b” and the upper ITO connected to the ground, the weakest region of the electric field is formed just above the microband “b”, and the cells move there by the repulsive force of negative dielectrophoresis (2nd patterning in Fig. B). At this time, if the cells adhered to the substrate via binding to molecules immobilized on the lower substrate, the cells remain on the gaps. Cells that do not interact with the molecules are expelled on the microband electrode. These procedures provide a rapid method for evaluating the expression of surface proteins that interact with molecules on the lower substrate in as little as 2 min; 60 sec in an adhesion and 60 sec in an expelling.The proposed method was applied to identify the differentiation status of HL60. Undifferentiated HL60 is known to express CD13 but not CD11b, while HL60 differentiated into neutrophils and monocytic lineages is known to up-regulate CD11b expression and down-regulate CD13 expression. In the device pretreated with anti-CD11b antibody, almost all HL60 formed well-defined linear patterns on the gaps between the microband electrodes at the 1st patterning by applying the ac voltage (100 kHz in frequency and 20 Vpp in voltage) for 60 seconds and aligned on the microband electrode “b” after 2nd patterning (day 0 in Fig. B). This result suggests that CD11b rarely expressed in HL60 at the state of undifferentiation. It was also confirmed that the expression levels of CD11b were extremely low by immunobiological fluorescent imaging. Conversely, HL60s differentiated with 1 μM all-trans retinoic acid (ATRA), which is well known as a inducer of HL60 to neutrophil, were randomly adhered on the gaps, and only a few of the adhered HL60s moved onto the microband electrodes after 2nd patterning (day 4 inf Fig. B). The increase of number of cells remaining on the gaps in differentiated HL60 represented the upregulation of CD11b expression. The expression of CD11b in differentiated cells cultured in ATRA-containing medium for 4 days was confirmed by immunostaining. There was a good correlation between the expression levels of CD11b obtained by the present method and those obtained by conventional fluorescent method. In addition, the expression of CD13 was investigated using the device pretreated with anti-CD13. In this case, the majority of undifferentiated cells adhered on the gap between the microband electrode and the differentiated cells were expelled from the gap onto the microband electrode “b”. Thus, using the repulsive force of n-DEP to expel cells unbound to the substrate, the differentiation status can be elucidated in as little as 120 seconds. Figure 1