Bacterial endotoxin (lipopolysaccharide; LPS) is known to alter systemic vascular permeability, but this effect is difficult to monitor and quantitate in vivo. The ocular vessels of the rabbit are particularly sensitive to LPS. Using a slit lamp equipped with a fluorophotometer, we have adapted a method to quantitate endotoxin-induced ocular vascular permeability by measuring the accumulation of fluorescein isothiocyanate-conjugated albumin into the anterior chamber of the eye. After intravenous administration of Salmonella typhimurim LPS, the anterior chamber fluorescence and blood fluorescence were measured at intervals of 15 min and 1 h, respectively, over 4 h. In controls, maximal fluorescence in the anterior chamber was 3.1 ± 0.8% of blood fluorescence. Doses of LPS as low as 0.25 μg/kg produced an ocular/serum fluorescence ratio of 17.6 ± 4.9. A dose of 2.5 μg of LPS per kg tended to produce a higher ratio (68.0 ± 7.1) than a larger dose of 50 μg/kg (30.5 ± 16.6). Permeability changes began within 30 min after LPS, and the rate of dye accumulation varied over time, with maximal leakage usually occurring 90 min after LPS, but occasionally occurring much later. Repeated doses produced tolerance. By conjugating albumin to rhodamine and utilizing a second filter with the slit lamp to measure accumulation of this dye, we demonstrated the persistence of marked permeability during a period when intraocular fluorescein isothiocyanate and albumin levels were relatively constant. This methodology indicates that extremely low doses of LPS induce ocular permeability changes and that neither the time course nor the dose response of this effect is linear. Ocular fluorophotometry is a sensitive, noninvasive technique to study the dynamics and pharmacology of LPS-induced permeability changes.