In 2010 there were 60.5 million people with glaucoma worldwide and this number is expected to increase to 79.6 million by 2020. In most glaucoma patients medical therapy consists of topical eyedrops. However, administration and compliance are often problematic. Surgical and laser treatment for glaucoma is traumatic for the patients, involving high costs and might require repetition of the procedure. Therefore the aim of the present work was the development of drug-eluting biodegradable implants designed to provide a localized and long-term (6 months to 2 years) sustained release of the drug that can be used in the treatment of glaucoma. The implant should degrade within the site of implantation, eliminating the need for further surgery. The implant should be introduced in the ophthalmologist o ce, under local anaesthesia, with high implications in patient recovery and costs. By delivering controlled amounts of drug the implant would be pharmacologically e cient and increase patient compliance. Several polymer processing and drug loading techniques were used in order to prepare controlled drug delivery systems (CDDS) for intraocular application. The chosen materials were poly("-caprolactone), PCL, because of its slow degradation, which makes it useful for long term delivery and poly(oxyethylene-b-oxypropyleneb-oxyethylene), Lu, because of its release modulation capacity. Moreover, they are both commercially available, inexpensive and well characterised polymers. Three drugs were incorporated: timolol maleate, acetazolamide and dorzolamide hydrochloride. They are all agents that can decrease intraocular pressure (IOP) in open-angle glaucoma. Two types of devices were prepared: monolithic (sponges, bers, disks) and hybrid (disks), using supercritical solvent impregnation (SSI), electrospinning, melt compression and solvent casting. PCL blends were successfully impregnated with timolol maleate using a SSI technique. SSI e ciency results suggested that the best impregnating conditions were obtained when a cosolvent was used and when speci c drug-polymer interactions occurred as a consequence of di erent chemical structures due to polymer blending. Pressure can be either a favourable factor, when there was enough drug a nity for the polymers, or an unfavourable factor, when weaker bonding was involved. Drug loading, heterogeneous/homogeneous dispersion of drug inside the matrix, hydrophilicity and crystallinity all in uenced the drug release. The in vitro release results suggested that a sustained drug release rate can be obtained by changing the SSI operational conditions and by modulating the composition of blends, as a mean to control crystallinity, hydrophilicity and drug a nity for the polymer matrix. Bicomponent bers of PCL and Lu were obtained by electrospinning. Acetazolamide and timolol maleate were loaded in the bers in di erent concentrations (below and above the drug solubility limit in polymer) in order to determine the e ect of drug solubility in polymer, drug state, drug loading and ber composition on ber morphology, drug distribution and drug release kinetics. High loadings bers (with drug in crystalline form) showed higher burst and faster release than low drug content bers, indicating that the release was more sustained when the drug was encapsulated inside the bers in an amorphous form. Moreover, timolol maleate was released faster than acetazolamide, indicating that drug solubility in polymer in uences the partition of drug between polymer and elution medium, while ber composition also controlled drug release. At low loadings total release was not achieved, suggesting that drug remained trapped in the bers. The modelling of release data implied a three stage release mechanism: a dissolution stage, a desorption and subsequent di usion through water lled pores, followed by polymer degradation control. Dorzolamide loaded disks (hybrid device) were prepared by solvent-casting of PCL/Lu and subsequent coating with PCL solution. By blending, crystallinity, water uptake and mass loss were modi ed relative to the pure polymers. Burst was diminished by coating the disks with a PCL shell. All samples presented burst release except PCL-coated samples that showed controlled release during 18 days. For PCL-coated samples, barrier-control of di usion coupled with partition control from the core slowed down the release, while for 50/50 Lu/PCL-coated samples, the enhancement in porosity of the core diminished partition-control of drug release. Non-linear regression analysis suggested that a degradation model fully described the release curve considering a triphasic release mechanism: the instantaneous di usion (burst), di usion and polymer degradation stages. MTT test indicated that the materials were not cytotoxic for corneal endothelial cells. A good in vitro{in vivo correlation was obtained, with similar amounts of drug released in vitro and in vivo. The decrease in IOP was similar to that obtained by dorzolamide eyedrop instillation. Implantable monolithic disks for glaucoma treatment were prepared by blending PCL, Lu and dorzolamide. Their in vivo performance was assessed by their capacity to decrease IOP in normotensive and hypertensive eyes. Drug mapping showed that release was complete from blend disks and the low molecular weight (MW) PCL after 1 month in vivo. The high MW PCL showed non-cumulative release rates above the therapeutic level during 3 months in vitro. In vivo, the brous capsule formation around the implant controlled the drug release, working as a barrier membrane. Histologic analysis showed normal foreign body reaction response to the implants. In hypertensive eyes, the most sustained decrease was shown by the high MW PCL. The blending o ers the possibility to manipulate the release rate and the amount of released drug in order to prepare devices tailored to the patients needs. The long term degradation of all the prepared constructs ( lms, bers, sponges and disks) was studied. The in uence on degradation rate of several factors (construct type, crystallinity, MW, drug presence, blending) was assessed through water uptake, mass loss, crystallinity and MW evaluation. The degradation rate was higher for blends than for PCL and it was similar between di erent types of blends. The low MW disks had a degradation rate that was lower by one order of magnitude than high MW constructs. Porosity was shown to be a very important factor because at initial stage (or initial porosity), it will enhance water uptake and degradation, while at a later stage (or developed porosity), it will decrease degradation rate because of diminished autocatalytic e ects. High initial porosity produced an acceleration of degradation for sponges, bers and lms when compared to disks, while developed porosity reduced degradation for drug-loaded disks when compared to disks without drug. Modelling of the experimental data suggested that the contribution of surface e ects was as signi cant as autocatalytic e ects in overall bulk degradation. This work has revealed some insights into possible polymer processing and drug loading techniques for the preparation of CDDS for intraocular delivery. It also presented some results regarding the preliminary pre-clinical evaluation of PCL-based implants. In vivo, the drug-eluting implants were able to reduce IOP in an animal model of glaucoma. Nevertheless, another extremely important issue has to be addressed: patient compliance. Patience compliance is of extreme importance especially in the therapy of chronic diseases because patients have to keep up constantly with their pharmacological regimen. The superiority of CDDS relative to conventional therapy has to be proven in long term compliance studies because this is one of the main reasons of developing CDDS therapy in the rst place.