Due to limited aqueous solubility of dexamethasone, a water-soluble cyclodextrin-containing dexamethasone complex was prepared to increase the topical bioavailability. In this paper, the difference in corneal and scleral penetration between cyclodextrin-containing eye drop and TobraDex ex vivo in NZ white rabbit was compared. At first, 6 healthy rabbits were sacrificed by overdosing with anesthetics to collect their corneas and scleras. And then, Franz diffusion pool was used to evaluate the permeability of the two formulations by calculating the cumulative permeation amounts (Qn). At last, concentrations of dexamethasone and tobramycin were determined using a validated LC-MS/MS method. The results showed that for cornea, Qn of dexamethasone in cyclodextrin-containing eye drops and TobraDex was 7.63μg/cm2 and 19.15μg/cm2 (p < 0.05), resp., and Qn of tobramycin was 45.27μg/cm2 and 322.28μg/cm2 (p < 0.05), resp. For sclera, Qn of dexamethasone of these two formulations was 132.44μg/cm2 and 30.45μg/cm2 (p < 0.05) and Qn of tobramycin was 964.38 and μg/cm2 1107.07μg/cm2 (p > 0.05), resp. Qn of both dexamethasone and tobramycin in cyclodextrin-containing eye drop was significantly lower than those in TobraDex. However, Qn of dexamethasone in cyclodextrin-containing eye drop was significantly higher in sclera, while tobramycin level was the same. So it is suggested that development of cyclodextrin-containing eye drop should have a potential advantage via trans-scleral administration such as sub-conjunctiva, sub-tenon and retrobulbar injection.