3D cells are encapsulated in the thick matreigel, drugs can not penetrate the matrigel evenly, resulting in the gradient effect. The gradient effect would affects the accuracy of the 3D cell-based biosensor response to drug screening. Traditional 3D cell impedance sensors is difficult to measure the gradient of 3D cell drug permeation effectively. In this study, a high-throughput 3D parallel impedance chip has been constructed for real-time drug gradient effect monitoring. Adenocarcinomic human alveolar basal epithelial cell lines (A549) are encapsulated in matrigel to construct 3D lung cancer model. The system consist of 3D parallel impedance chip with 16 vertically aligned gold electrodes, a homemade detecting instrument and a PC software based on LabVIEW. Computational finite element (FE) analysis was used to explore the interactions of adjacent electrodes. In our previous study, cell proliferation or apoptosis in the matrigel would cause variation of cell density, which shown to be inverse proportional to the impedance magnitude of the cell/matrigel entire construct. By using this system, long-term, non-invasive, real-time monitoring of cell proliferation and viability can be achieved. Moreover, after the cell proliferation reached to plateau phase, 3D cell culture system was dosed with different kind of anti-cancer drugs under diverse concentration and the viability of 3D cells at different spatial location sites can be monitored. The experiment results show that drug permeability gradient can be detected by our sensing system. This high-throughput impedance sensing system has a high potential to develop a powerful analytical platform for cancer research and drug screening. Key words: 3D cell culture, 3D parallel impedance chip, permeability gradient, cell proliferation, impedance measurement,drug screening Introduction Conventional 2D cell models have been widely introduced in various researches, but it is too simplified to truly mimic the microenvironment in vivo. Besides, 2D cell model often succumbs to treatment and drugs appear to be very effective. Recently, 3D cancer cell models have attracted increasing interests in the study of drug screening and assessment. Researchers have found that 3D cell model can form a dynamic spatial gradient of soluble factors, which affect cell migration, intercellular communication and differentiation in a more physiological manner. Compared to 2D cell model, 3D cell model are often more resistant to treatment and being better predictors of in vivo drug responses. However, there exist practical problems such as drug release, drug permeation due to the special property of 3D cell model. Traditional single-channel 3D ECIS sensing chips can real-time monitor the whole 3D hydrogel mixed with cells entirely, but have no way to distinguish the spatial distribution of cells at different spatial location sites in larger scaffolds. Confocal fluorescence microscope is used as a gold standard for analysis of 3D cells viability and differentiation. Although confocal microscope provides accurate morphological characterization for 3D cell, it is still a labor-intensive, time-consuming and endpoint analysis method ,fails to meet the high-throughput demand of drug screening. We proposes a sensor to evaluate drug permeability gradient effect in real time. Method The system was constructed for 3D cancer cell culture, assay and drug permeability evaluation. Matrigel incorporated with A549 cells seeds into cell culture plate, cell viability can be monitored by our 3D ECIS chips at the same time. Cell index (CI) is proposed as equation CI=|Zt-Z0|/Z0 to acquire a normalized results, where Z0 is the background impedance, Zt is the measured impedance at a certain moment. Our high-throughput impedance sensing chip consists of 16 pairs of electrodes, which are 0.5×0.5mm square electrodes. A549 cells suspended in cell culture medium(RPMI 1640 with 10% FBS) were mixed in matrigel and Cross-linked in a 37 °C humidified incubator with 5% CO2 for 30 minutes. Drug test conducted by cisplatin,pemetrexed and gemcitabine. Results and Conclusions When an external stimulant electrical signal was imposed to the entire structure, gap junctions of cells electrically connected, causing the increasement of conductivity. As the number of cells in the scaffold increases, the number of gap junction increases, causing a further increase in conductivity of the structure (Fig.1A).The system consists of microcontroller, direct digital synthesizer, amplification and filter module and AD sampling module. 16 vertically aligned pair of electrodes were embedded at the opposite sidewalls of the sensing chip for the on-site impedance measurement(Fig.1B,Fig.1C). The impedance-frequency plot of the 3D cell construct is showed in Fig 1D. At the frequency of 10 to 1000 kHz, the impedance magnitude presents significant difference. Based on hardware cost considerations, we choose 10KHz as the simulation source frequency. Stability test was also executed (Fig.1E). A549 cells were selected as 3D cancer cell model. Fig. 1F&Fig.1G showed the first and sixth day morphology and viability of 3D cultured A549 cells separately.Fig.1H showed the cell growth curves of 16 channels in the density of 10000 cells/well.Fig.1I shows the drug permeability gradient effect. Figure 1