The permeability of the blood-retinal and blood-aqueous barriers to fluorescein (F) and the rate of aqueous flow can be estimated by measurements of F in the vitreous, aqueous, and plasma after systemic administration. F is commonly measured by fluorescence, but fluorescein glucuronide (FG), a metabolite of F, also fluoresces. To assess the influence of FG on the quantitation of F by fluorescence, we studied the pharmacokinetics of F and FG for 38 hr in the plasma of five normal subjects given 14 mg/kg of sodium fluorescein intravenously. The plasma and the plasma ultrafiltrate were measured by fluorescence and by high performance liquid chromatography. In our fluorophotometer, FG was 0.124 times as fluorescent as F. F was rapidly converted to FG, and within 10 min the concentration of unbound FG exceeded that of unbound F. The terminal half-lives of F and FG in the plasma ultrafiltrate were 23.5 and 264 min, respectively, so that FG contributed almost all of the plasma fluorescence after 4-5 hr. Because FG was less bound in the plasma than F, the ratio of the fluorescence of the plasma ultrafiltrate to that of the plasma increased with time. The greatest proportion of the total F available to penetrate into the ocular compartments occurred shortly after injection. We concluded that FG is an important contributor to the fluorescence of the plasma ultrafiltrate after intravenous injection and that accurate quantitation of physiologic parameters calculated from the plasma F requires taking this factor into account.