Protein crosslinking photosensitized by rose Bengal (RB2−) has multiple medical applications and understanding the photosensitization mechanism can improve treatment effectiveness. To this end, we investigated the photochemical efficiencies of monomeric RB2− (RBM2−) and dimeric RB2− (RBD2−) and the optimal pH for anaerobic RB2− photosensitization in cornea. Absorption spectra and dynamic light scattering (DLS) measurements were used to estimate the fractions of RBM2− and RBD2−. RB2− self‐photosensitized bleaching was used to evaluate the photoactivity of RBM2− and RBD2−. The pH dependence of anaerobic RB2− photosensitization was evaluated in ex vivo rabbit corneas. The 549 nm/515 nm absorption ratio indicated that concentrations > 0.10 mm RB contained RBD2−. Results from DLS gave estimated mean diameters for RBM2− and RBD2− of 0.70 ± 0.02 nm and 1.75 ± 0.13 nm, respectively, and indicated that 1 mm RB2− contained equal fractions of RBM2− and RBD2−. Quantum yields for RB2− bleaching were not influenced by RBD2− in RB2− solutions although accounting for RB2− concentration effects on the reaction kinetics demonstrated that RBD2− is not a photosensitizer. Optimal anaerobic photosensitization occurred at pH 8.5 for solutions containing 200 mm Arg. These results suggest potential approaches to optimizing RBM2−‐photosensitized protein crosslinking in tissues. Rose Bengal (RB2−)‐photosensitized crosslinking of protein in tissues has many applications in medicine. To improve the effectiveness of these treatments, we investigated the photochemical efficiencies of monomeric RB2− and dimeric RB2−. Dynamic light scattering (DLS) indicated that the treatment solution used for medical applications (1 mm RB2− in PBS) contains a 1:1 mixture of RB monomers and dimers. Studies of the singlet oxygen‐mediated RB2− self‐photosensitized bleaching indicated that the RB dimer is not a photosensitizer, suggesting that the efficiency of tissue protein crosslinking may be enhanced by blocking RB2− dimerization.