Purpose: To compare the effects between intravitreal and periocular injections of adenoviral vectored pigment epithelium-derived factor (AdPEDF) on suppressing established choroidal neovascularization (CNV) in rats. Methods: Sixty-eight female BN rats (136 eyes) aged 6–8 weeks were involved in this study. The CNV animal models were created with laser photocoagulation. After CNV, 16 rats underwent intravitreal injection with AdPEDF 1 µl (group A), 16 rats received intravitreal injection with control vector (AdNull) 1 µl (group B), 16 rats had periocular injection with AdPEDF 1 µl (group C), and 16 rats had periocular injection with AdNull 1 µl (group D). The effects between intravitreal and periocular AdPEDF injections on suppressing established CNV in rats were compared and evaluated by fundus fluorescein angiography (FFA), maximal thickness of CNV, histopathology and transferase-mediated uridine nick end labeling (TUNEL) staining 3, 7, 14 and 28 days after treatment. Results: (1) There were no significant changes in leakage in groups A and C 3 days after injection compared with that before injection seen by FFA (p > 0.05). The leakages in groups A and C decreased significantly 7 days after injection compared with that before injection (p < 0.05). (2) There was no significant difference in the incidence of CNV between groups A and B, as well as groups C and D 3 days after injection (p > 0.05). The incidence of CNV decreased significantly in group A compared with that in group B 7, 14 and 28 days after injection (p < 0.01). CNV retained fibrovascular proliferation in groups B and D 7 days after injection. (3) The maximal thickness of CNV in groups A and C diminished significantly compared with that in the control group after injection, and it still diminished with time (p < 0.05). There was no significant difference in maximal thickness of CNV between 14 and 28 days after injection of AdPEDF (p > 0.05). The maximal thickness of CNV in group A was larger than that in group C 3 days after injection (p < 0.05), yet it was smaller than that in group C 14 and 28 days after injection (p < 0.05). (4) Histopathologically, a great deal of CNV was shown 3 days after injection of AdPEDF or AdNull. CNV decreased significantly with lumen diminution 14 days after injection of AdPEDF. (5) TUNEL cells appeared in groups A and C 7, 14 and 28 days after injection, and there were no TUNEL cells in groups B and D. TUNEL cells were not seen in choroidal inherent vascular endothelial cells in all groups. Conclusion: Compared with the effect of periocular AdPEDF injection on suppressing established CNV in rats, the effect of intravitreal injection started slowly, but lasted longer. The effect appeared on day 7, reached the peak on day 14 and remained stable on day 28 after the treatment.