30 patients with corneal epithelial defect due to removal of corneal foreign bodies were randomly allocated to receive either chloramphenicol with continuous application of an eyepad or chloramphenicol without the eyepad. Almost all corneal defects were healed at 24 h, and all were healed by 48 h, with no statistically significant difference between the two groups. Discomfort at 24 h was greater in the eyepad group than in the control group. An eyepad seems to confer no benefit in healing and is uncomfortable.