The purpose of this study was to investigate the permeability and clinical significance of the blood-ocular barrier. A new technique using systemic, subtenon, and retrobulbar injections was applied to the study of intraocular penetration of labelled steroid compound. The study was carried out in normal rabbits and in animals with artificially induced intraventional inflammation. It is concluded that there is no blood-vitreous barrier, a vague concept in any case, in the anterior region of the vitreous where free diffusion was observed between the anterior and posterior segments of the eye. High concentration of the labelled steroids was found in the cornea with the peak of activity usually at 30 minutes. However, traces of activity were detectable for up to 4 hours. This indicates the need for a depot long term respiratory form of corticosteroid that would deliver a high concentration of the medication. Local subtenon's and retrobulbar injections resulted in a relatively higher ocular I-125 Cortisol concentration than obtained by systemic route. It is concluded therefore that a high concentration of steroids applied locally will give better and less deletrious effects than systemic administration. Subtenon injection resulted in higher activity in various ocular tissues especially aqueous and vitreous. However, the I-125 Cortisol was rapidly cleared regardless of the route of administration.