Although some investigations have demonstrated the ability of inflammatory mediators, including vasopermeability and chemotactic factors, to induce acute inflammatory reactions in vivo, little is known about the response of various elements of the anterior segment to the direct effects of inflammatory mediators. These studies were initiated to develop models for the investigation of inflammatory responses in this region of the eye. Acute inflammatory reactions were induced within the rabbit anterior chamber by intracameral injection of 50 microliters of various inflammatory mediators and were evaluated by clinical grade, leukocyte influx into the aqueous humor, and morphologic changes in the corneal endothelium. Peak responses were recorded following injection of 10(-4) M formyl-methionyl-leucyl-phenylalanine (fMLP); 5 ED50 C5fr; 0.5 mg/ml C5; undiluted anti-red blood cell (RBC) serum; and 10(-5) M histamine. The number of leukocytes per milliliter of aqueous humor induced by each mediator was quantitated by comparison with the number of leukocytes induced by buffer instillation into a separate group of rabbits (mediator-induced influx/buffer-induced influx). Comparisons were made 24 hours after instillation of mediators. The results of these studies were as follows: buffer alone, 1.0; fMLP, 3.1 C5fr, 61.0; C5, 8.7; anti-RBC, 91.0; and histamine, 24.0. Clinical grades correlated well with these ratios. In addition, differences were noted when the time kinetics of acute responses induced by two different mediators (10(-4) M fMLP, a synthetic preformed chemotactic factor; and a 1:5 dilution of anti-RBC, which binds to vascular and corneal endothelial cells) were directly compared over 48 hours. Responses induced with fMLP peaked between 5 and 8 hours and resolved rapidly, whereas anti-RBC-induced responses peaked between 8 and 12 hours and resolved very slowly. Histopathologic analysis indicated that both fMLP and anti-RBC induced a similar sequence of changes in the corneal endothelium. Within 2-3 hours after instillation of either mediator, the endothelial cells exhibited prominent vacuolization/retraction phenomena. At the peak of leukocyte influx PMNs filled these vacuoles, then migrated back into the aqueous humor within several hours. Normal morphologic features were recovered following clearance of leukocytes from the anterior chamber. We believe that these models will be useful in identifying the roles of individual mediators in acute and chronic endocular inflammation and in the injury of corneal endothelium.