The use of microfluidic gradient generators to study cell chemotaxis has been demonstrated for a number of different cell-culture systems. However, because current gradient generators require fluid flow through the cell-culture chamber, these devices are not capable of culturing primary neurons, which are extremely sensitive to shear stress. Therefore, we have developed a novel, microfluidic gradient generator that allows the cell-culture chamber to remain static while applying a stable concentration gradient across the chamber. The device consists of two parallel channels connected by a series of small tunnels to a large cell-culture chamber. Cell-culture medium with and without supplemental chemotaxis cues is driven through the channels using a syringe pump, allowing Brownian diffusion to carry the chemotaxis cue through the small tunnels and into the cell-culture chamber. A stable concentration gradient (as viewed using fluorescent microscopy) is quickly developed and maintained indefinitely while the system is at equilibrium. The device is capable of producing stable gradients of both small molecules as well as proteins. Multiple gradient generators and cell-culture chambers can easily be multi-plexed into a single device, allowing for optimum flexibility and reproducibility. To date, we have successfully cultured primary rat hippocampal neurons and Xenopus spinal cord neurons in the device. We observe good cell viability over four days of culture, a critical time period during which neurons mature and become polarized. The immature neurons in our devices undergo normal polarization and develop an axonal process that sprouts from the cell body. Using our devices, we can now measure the minimum concentration gradient required to guide the path-finding of the axon in response to known guidance cues. The device should also prove useful in identifying novel guidance cues that regulate this developmental process and guide axonal movement. Another feature of the device is the ability to test multiple environmental cues simultaneously to discern potential signaling crosstalk. Future studies will evaluate opposing and reinforcing gradients of guidance cues to identify combinations that optimize the distance over which the axon can be guided. Beyond having an immediate impact on our understanding of developmental neurobiology, our long-term goal is to translate this information to the regeneration and guidance of injured axons. Our design utilizes a novel approach to incorporating shear-sensitive cells within a microdevice and is therefore ideally suited for culture of other "difficult" cell types such as stem cells.