Aims: To evaluate and compare the inhibitory effects of topical and subconjunctival bevacizumab on corneal neovascularization in a rat model. Methods: Twenty corneas of 20 rats were chemically cauterized with silver nitrate sticks. Animals were randomized into four groups: a control group that received only topical artificial tear drops twice daily, a subconjunctival injection group that received 1.25 mg (0.05 ml) of bevacizumab on the 1st, 4th, and 7th day, and two topical bevacizumab groups that received instillation of 4 or 12.5 mg/ml bevacizumab twice daily. Digital photographs of the cornea were taken and analyzed using an image analysis software program. On the 10th day, corneas were excised and examined histologically. Results: The mean percentage of the vascularized corneal area (%) in the control group was 63.32 ± 13.10 (mean ± SD), compared with 30.22 ± 15.73 in the subconjunctival injection group, 26.76 ± 10.23 in the 4-mg/ml topical group, and 25.52 ± 12.45 in the 12.5-mg/ml group. The differences between the control and each treatment group were significant (all p < 0.01). Further, histological examination revealed that each treatment group had fewer vessels than the control group (all p < 0.01). Conclusions: Both subconjunctival injection and topical use of bevacizumab are effective and safe in controlling corneal neovascularization.