||The development of devices to electrically interact with the brain is a challenging task that could potentially restore motion to paralyzed patients and sight to those with profound blindness. Neural engineers have designed many types of microelectrode arrays (MEAs) with this challenge in mind. These MEAs can be implanted into brain tissue to both record neural signals and electrically stimulate neurons with high selectivity and spatial resolution. Implanted MEAs have allowed patients to control of a variety of prosthetic devices in clinical trials, but the longevity of such motor prostheses is limited to a few years. Performance decreases over time as MEAs lose the ability to record neuronal signals, preventing their widespread clinical use. Microstimulation via intracortical MEAs has also not achieved broad clinical implementation. While microstimulation for the restoration of vision is promising, human clinical trials are needed. Chronic in vivo functionality assays in model systems will provide key insight to facilitate such trials. There are three goals that may help address insufficient MEA longevity, as well as provide insight on microstimulation functionality. First, thorough characterizations of how performance decreases over time, both with and without stimulation, will be needed. Next, factors that affect the chronic performance of microstimulating MEAs must be further investigated. Finally, intervention strategies can be designed to mitigate these factors and improve long term MEA performance. This dissertation takes steps towards meeting these goals by means of three studies. First, the chronic performance of intracortically implanted recording and stimulating MEAs is examined. It is found that while performance of implanted MEAs in feline cortex is dynamic, catastrophic device failure does not occur with microstimulation. Next, a variety of factors that affect microstimulation studies are investigated. It is found that many factors, including device iv damage, anesthesia depth, the application of microstimulation, and the use of impedance as a reporter play a role in observations of performance variability. Finally, a promising intervention strategy, a carbon nanotube coating, is chronically tested in vivo, indicating that carbon nanotubes do not cause catastrophic device failure and may impart benefits to future generations of MEAs.