Neuroprostheses were developed for a variety of applications using unique microfabrication methods. A technology platform that enables novel microelectrode array designs and microfluidic channels is described. Experiments demonstrating the in vivo utility and unique capabilities of these neuroprostheses were performed. Despite their demonstrated effectiveness, the long term use of microscale neuroprosthetic devices is compromised by the post-implantation tissue reaction which forms around the device. This tissue reaction consists of glial cells on the same size scale as microelectrodes and microfluidic channels. In order to quantify the tissue reaction a new analytical method using Electrical Impedance Spectroscopy was developed called Peak Resistance Frequency analysis. A lumped parameter model describing the microelectrode-tissue interface was elaborated which identified and isolated the electrical characteristics of the different interface components. The model accurately predicted how the post-implantation tissue reaction increases the electrical impedance of the interface as demonstrated by in vivo experiments. The major goal of this work was to determine whether the tissue reaction to an implanted neuroprosthesis could be reduced using a controlled drug release mechanism and to measure the effect quantitatively. It was hypothesized that highly localized delivery for several days of an anti-inflammatory drug around the implant would disrupt the tissue reaction, thereby decreasing the electrical impedance of the microelectrode-tissue interface. Controlled release drug coatings were designed to deliver dexamethasone-loaded nanoparticles in the immediate region surrounding the implant. Drug-eluting and control probes were evaluated in the rat cortex. The electrical resistance of the encapsulation tissue was isolated and its progression as a function of post-implantation time was compared using Peak Resistance Frequency analysis and immunohistochemistry. These in vivo experiments demonstrated a decreased tissue reaction for drug-eluting probes and the effect was sustained chronically, therefore proving the hypothesis. These techniques offer a solution to the problem of tissue reaction around microscale neuroprostheses and an improvement in neural stimulation and recording capability.