Chronic diseases can often be managed by constantly delivering therapeutic amounts of drug for prolonged periods. A controlled release for extended duration would replace the need for multiple and frequent dosing. Local drug release would provide added benefit as a lower dose of drug at the target site will be needed as opposed to higher doses required by whole body administration. This would provide maximum efficacy with minimum side effects. Nonetheless, a problem with the known implantable drug delivery devices is that the delivery rate cannot be controlled, which leads to drug being released in an unpredictable pattern resulting in poor therapeutic management of patients. This dissertation is the result of development of an implantable drug delivery system that is capable of long-term zero order local release of drugs. The device can be optimized to deliver any pharmaceutical agent for any time period up to several years maintaining a controlled and desired rate. Initially significant efforts were dedicated to the characterization, biocompatibility, and loading capacity of nanoporous metal surfaces for controlled release of drugs. The physical characterization of the nanoporous wafers using Scanning electron microscropy (SEM) and atomic force microscopy techniques (AFM) yielded 3.55 x 10⁴ nm³ of pore volume / μm² of wafer surface. In vitro drug release study using 2 - octyl cyanoacrylate and methyl orange as the polymer-drug matrix was conducted and after 7 days, 88.1 ± 5.0 % drug was released. However, the initial goal to achieve zero order drug release rates for long periods of time was not achieved. The search for a better delivery system led to the design of a perforated microtube. The delivery system was designed and appropriate dimensions for the device size and hole size were estimated. Polyimide microtubes in different sizes (125-1000 μm) were used. Micro holes with dimensions ranging from 20-600 μm were fabricated on these tubes using photolithography, laser drilling, or manual drilling procedures. Small molecules such as crystal violet, prednisolone, and ethinyl estradiol were successfully loaded inside the tubes in powder or solution using manual filling or capillary filling methods. A drug loading of 0.05 – 5.40 mg was achieved depending on the tube size and the drug filling method used. The delivery system in different dimensions was characterized by performing in vitro release studies in phosphate buffered saline (pH 7.1-7.4) and in vitreous humor from the rabbit’s eye at 37.0 ± 1.0°C for up to four weeks. The number of holes was varied between 1 and 3. The tubes were loaded with crystal violet (CV) and ethinyl estradiol (EE). Linear release rates with R²>0.9900 were obtained for all groups with CV and EE. Release rates of 7.8±2.5, 16.2±5.5, and 22.5±6.0 ng/day for CV and 30.1±5.8 ng/day for EE were obtained for small tubes (30 μm hole diameter; 125 μm tube diameter). For large tubes (362-542 μm hole diameter; 1000 μm tube diameter), a release rate of 10.8±4.1, 15.8±4.8 and 22.1±6.7 μg/day was observed in vitro in PBS and a release rate of 5.8±1.8 μg/day was observed ex vivo in vitreous humor. The delivery system was also evaluated for its ability to produce a biologically significant amounts in cells stably transfected with an estrogen receptor/luciferase construct (T47D-KBluc cells). These cells are engineered to produce a constant luminescent signal in proportion to drug exposure. The average luminescence of 1144.8±153.8 and 1219.9±127.7 RLU/day, (RLU = Relative Luminescence Units), yet again indicating the capability of the device for long-term zero order release. The polyimide device was characterized for biocompatibility. An automated goniometer was used to determine the contact angle for the device, which was found to be 63.7±3.7degreees indicating that it is hydrophilic and favors cell attachment. In addition, after 72 h incubation with mammalian cells (RAW 267.4), a high cell distribution was observed on the device’s surface. The polyimide tubes were also investigated for any signs of inflammation using inflammatory markers, TNF-α and IL-1β. No significant levels of either TNF-α or IL-1β were detected in polyimide device. The results indicated that polyimide tubes were biocompatible and did not produce an inflammatory response. / text