The goal of this study was to create pre-formed neural spheroid networks (NSN) on a polydimethyl siloxane (PDMS) concave microwell array for eventual implantation into the rat brain. Recent studies have shown that stem cells have great potential in treating various neurological insults of the central nervous system, ranging from traumatic brain and spinal cord injury, to neurodegenerative disorders. However, the use of stem cell lines in research are controversial due to the method of obtaining cells, in their formation of teratomas and degeneration into cancer cells, their non-specific differentiation, and lastly in their inability to control the location of neural connections. A novel approach to address this issue utilizes pre-formed neural networks consisting of neural spheroids on polymer scaffolds for the implantation into the rat brain. Yet, it was observed that the cylindrical shape of the wells hindered the transfer process. This study aimed to overcome the lack of neural spheroid network detachment by utilizing concave well structures, using a simple method developed in this laboratory. Primary neurons were isolated from pregnant Sprague Dawley rats at 16 ~ 17 days of gestation. Isolated neurons were cultured in PDMS wells with a concave structure and interconnected by rounded micro channels. It was reported previously that a concave structure enabled an easier and more efficient formation of spheroids, not to mention the ease in extraction of spheroid cells. Various studies have demonstrated the effectiveness of guidance channels in promoting neurite growth. Therefore, micro channels were integrated in the micro array design, and served as a guidance conduit to enhance neurite growth, and by association, spheroid interconnection. The primary neurons formed a spheroid structure after 3 days, upon which they began to sprout new neurites. By day 8, neurite connections peaked. Spheroid diameter underwent an initial decrease then stabilized on day 2. Various well diameters (300~700 um) and channel lengths (1.5 x diameter ~ 3 x diameter) were evaluated, with a 300 um well diameter and 450 um center-to-center channel length found to be optimal. The completed network was assessed for interconnection using calcium imaging and showed coordinated calcium signals between the neural spheroids. The network was then successfully transferred to a collagen matrigel and cultured for a week. The methodology showed an improvement in the transfer of networks, with about a 90% extraction rate. The viability of the NSN on the matrigel was assessed using a Live/Dead assay, and cells were found to have greater than 95% viability. The optimal hydrophilicity was determined for neurite extension and transfer of NSNs onto the matrigel. It was found that an incubation time between 4~6 hours was optimal. Future studies will involve the implantation of the NSN into the rat brain. Additionally, the use of neural progenitor and stem cell lines may provide an autologous source of cells which are immunocompatible with the host. In particular, marrow stromal cells are interesting in that they may also address the ethical concerns. A long term goal is to refine the methodology and apply this research to enable studies in the treatment of patients suffering from spinal cord injury and other neurodegenerative disorders.