Rabbit eyes were subjected to severe alkali burns (35 sec, 12 mm, 1N sodium hydroxide). In one experiment, rabbits in the treated group received a daily subcutaneous injection of neutralized ascorbic acid solution (0.5 gm/kg body weight), while control animals received no treatment. At the termination of the experiment (30 days), 11 of 16 eyes (68.8%) in the control group had ulcerated or formed descemetoceles, and in the experimental (treated) group, 15 of 20 eyes (75%) had ulcerated, formed descemetoceles, or perforated. In a second experiment, burned rabbits received topical 10% ascorbic acid while control eyes were given the vehicle only. At the termination of the experiment (34 days), 16 of 20 eyes (80%) in the control group had ulcerated or perforated, compared to five of 18 eyes (27.8%) in the ascorbate treated groups. The failure of systemic administration of ascorbic acid to prevent corneal ulceration could be explained on the basis of inadequate penetration of ascorbic acid into the anterior segment of severely burned rabbit eyes. On the other hand, immediate topical treatment of identically burned rabbit eyes achieved greatly elevated aqueous humor ascorbate levels and provided substantial protection from corneal ulceration and perforation.