The results are reported of a fully controlled randomized double-blind clincial trial of adenine arabinoside and idoxuridine ointment in sixty patients with herpetic ulceration of the cornea. Although both antivirals showed a trend towards superiority over placebo, the therapeutic effect did not reach statistical significance in spite of the known efficacy in laboratory animals. Further studies in rabbits are reported; these indicate that systemic immunity may play a role in combating virus proliferation in recurrent disease, and it is considered this disguises the efficacy of topical antiviral therapy in clinical trials, thus necessitating an estimated requirement for approximately fifty patients per treatment group to obtain significant effects. It is concluded that an antiviral is valuable in the treatment of ulcerative herpetic keratitis, particularly in primary disease and in the presence of systemic and local immunosuppression after the use of topical adrenocorticosteroid. In recurrent disease, where a trigger factor is known, experience has shown that therapy can be profitably administered before the onset of clinical disease.