Despite differences in the processes leading to tissue damage, the ocular irritation response to various surfactants, two concentration s of an acid and an alkali, and an acetone, alcohol, aromatic amine, and aldehyde has been shown to depend on the extent of initial injury. The purpose of this study was to assess the extent to which this fundamenta l relationship exists for bleaching agents in the rabbit low-volume eye test. Ten l l of sodium perborate monohydrat e (NaBO 3 ), sodium hypochlorite (NaOCl), 10% hydrogen peroxide (H 2 O 2 ), and 15% H 2 O 2 was applied directly to the cornea of the right eye of each rabbit. Macroscopic assessments for irritation were made 3 hours after dosing and periodically until 35 days. Light microscopic examinations were conducted on tissues obtained at 3 hr and on 1, 3, and 35 days. In vivo confocal microscopy (CM) and measurements of dead corneal epithelial cells and keratocytes at 3 hours and 1 day were used to characterize quantitatively initial corneal injury, while in vivo CM performed at 3 hours and 1, 3, 7, 14, and 35 days was used to characterize quantitatively the corneal changes over time. The changes with NaBO 3 and NaOCl were consistent with mild irritancy. For both, corneal injury was limited to the epithelium and super cial stroma. The changes with 10% H 2 O 2 and 15% H 2 O 2 were consistent with severe irritation. Both concentrations affected the epithelium and deep stroma, with 15% H 2 O 2 also at times affecting the endothelium. However, unlike other irritants previously studied, with 10% H 2 O 2 and 15% H 2 O 2 there was an incongruit y between the extent of epithelial and stromal injury, with stromal injury being more extensive than epithelial injury. A similar, although less dramatic, effect was observed with NaBO 3 . Additionally, there was still signi cant keratocyte loss at 35 days with 10% H 2 O 2 and 15% H 2 O 2 even though the eyes at times were considered to be macroscopicall y normal. These observations highlight the need to include both epithelial and stromal component s in an ex vivo or in vitro alternative assay. In conclusion, these results continue to support and extend our hypothesis that ocular irritation is principally de ned by the extent of initial injury despite clear differences in the means by which irritants cause tissue damage. Importantly, we have identi ed unique differences in the ocular injury and responses occurring with bleaching agents that are important to consider in the development and validation of alternative ocular irritation tests to characterize a broad range of materials differing in type and irritancy.