PURPOSE: The aim of this study was to measure and compare the effect of topical insulin (0.5 units, 4 times per day) versus artificial tears (Vismed, sodium hyaluronate 0.18%, 4 times per day) for the healing of postoperative corneal epithelial defects induced during vitreoretinal surgery in diabetic patients. METHODS: This is a double-blind randomized controlled hospital-based study involving diabetic patients with postoperative corneal epithelial defects after vitreoretinal surgery. Diabetic patients were randomized into 2 different groups and received either 0.5 units of topical insulin (DTI) or artificial tears (Vismed, sodium hyaluronate 0.18%; DAT). The primary outcome measured was the rate of corneal epithelial wound healing (mm 2 /h) over a preset interval and time from baseline to minimum size of epithelial defect on fluorescein-stained anterior segment digital camera photography. The secondary outcome measured was the safety of topical insulin 0.5 units and artificial tears (Vismed, sodium hyaluronate 0.18%). Patients were followed up until 3 months postoperation. RESULTS: A total of 38 eyes from 38 patients undergoing intraoperative corneal debridement during vitreoretinal surgery with resultant epithelial defects (19 eyes per group) were analyzed. DTI was observed to have a significantly higher healing rate compared with the DAT group at rates over 36 hours ( P = 0.010), 48 hours ( P = 0.009), and 144 hours ( P = 0.009). The rate from baseline to closure was observed to be significantly higher in the DTI group (1.20 ± 0.29) (mm 2 /h) compared with the DAT group (0.78 ± 0.20) (mm 2 /h) as well ( P < 0.001). No adverse effect of topical insulin and artificial tears was reported. CONCLUSIONS: Topical insulin (0.5 units, 4 times per day) is more effective compared with artificial tears (Vismed, sodium hyaluronate 0.18%, 4 times per day) for the healing of postoperative corneal epithelial defects induced during vitreoretinal surgery in diabetic patients, without any adverse events.