Objective To explore the short-term efficacy and safety of intra-arterial thrombolysis (IAT) in the treatment of retinal artery occlusion (RAO) with the assistance of the rescue green channel in the eye stroke center. Methods A prospective, interventional, single-center study. Thirty-eight eyes from 38 RAO patients who received IAT treatment in Guangdong Provincial People’s Hospital were enrolled. All the patients were rescued via the green channel in our eye stroke center. Data from comprehensive ocular examinations including best-corrected visual acuity (BCVA) and optical coherence tomography angiography (OCTA) were collected. BCVA was measured with Snellen chart and converted to the logarithmic minimum angle of resolution (logMAR) unit for statistical analysis. RTVue XR OCTA was used to measure vascular densities (VD) of the superficial capillary plexus (SCP), deep capillary plexus (DCP) and radial peripapillary capillary (RPC), and central retinal thickness (CRT). All RAO eyes attempted IAT treatment and 34 eyes were successful. Four eyes failed to complete IAT because of the occlusion of internal or common carotid arteries on the same side with the RAO eyes. Ocular examinations in post-operative 1-3 days were performed with the same devices and methods as those before surgery. Parameters measured before and after surgery include BCVA, VD of SCP, DCP, RPC, and CRT. Data of the green channel collected include the time intervals from onset of RAO to first presentation in local hospitals, and from onset of RAO to our eye stroke center. Comparisons of VD and CRT between the RAO eyes and contralateral healthy eyes were performed with independent samples Mann-Whitney U test; comparisons of VD and CRT in RAO eyes before and after IAT surgery were performed with paired samples Wilcoxon Rank Sum test. Results Among the 34 RAO patients who had successful IAT surgery, 18 (52.9%, 18/34) were males and 16 (47.1%, 16/34) were females; the mean age was (51.0±12.9) years old. There were 30 and 4 eyes diagnosed as central RAO and branch RAO respectively. The logMAR BCVA before and after IAT surgery was 2.52±0.61 and 2.18±0.85 respectively, and the difference was statistically significant (Z=－3.453, P=0.002). Before surgery, VD of SCP, DCP and RPC were significantly decreased and CRT was significantly increased in the affected eye compared with the contralateral healthy eyes, with the statistical significance (P＜0.001). Compared with those before surgery, the VD of SCP and DCP were significantly improved after surgery (Z=－2.523, －2.427; P=0.010, 0.014), while there was no difference in VD of RPC and CRT (Z=－1.448, －1.454; P=0.150, 0.159). The time interval between onset of RAO and first visit to the hospital was (6.56±6.73) hours; the time interval between onset of RAO and the arrival at our hospital was (24.11±19.90) hours. No cerebral stroke was observed in the early post-operative period and no cerebrocardiovascular events were observed later. he time interval between onset of RAO and the arrival at our hospital was (24.11±19.90) hours. No cerebral stroke was observed in the early post-operative period and no cerebrocardiovascular events were observed later. Conclusions The short-term efficacy and safety of IAT in the treatment of RAO were satisfactory. The rescue time window might be prolonged.