PURPOSE: To evaluate the effects of oxygen therapy on corneal edema after cataract surgery. SETTING: Imam Khomeini Hospital, Ahvaz, Iran. DESIGN: Randomized controlled trial. METHODS: Patients with severe corneal edema were randomized into 3 groups. Group 1 (control) received conventional therapy including topical sodium chloride, timolol, and betamethasone. Group 2 received the same therapy in addition to systemic normobaric oxygen at a flow rate of 10 L/min for 1 hour twice daily for 3 weeks. Group 3 received conventional therapy and transcorneal oxygen at a flow rate of 5 L/min for 1 hour twice daily for 3 weeks. Preoperative pachymetry and specular microscopy were performed. Pachymetry was performed 1, 3, 5, 7, 10, and 14 days postoperatively. At 1, 3 and 12 months, pachymetry and specular microscopy were performed. RESULTS: The study enrolled 45 patients. Preoperatively, there was no significant difference between the groups. Pachymetry was more than 1000 mu m 1 day postoperatively in all patients. The preoperative pachymetry was restored in 14 days in Group 3 only. After 1 year, the endothelial cell count (ECC) was 1603 cells/mm(2), 1693 cells/mm2, and 1911 cells/mm(2) with a loss of 37%, 32%, and 25% in Group 1, Group 2, and Group 3, respectively (P = .034, Groups 1 and 3). Group 3 had a higher ECC than the control group at 3 months (P = .037) and 1 year (P = .025). One patient in Group 1 and 1 patient in Group 2 developed bullous keratopathy. CONCLUSION: Transcorneal oxygen decreased corneal edema more rapidly than conventional and systemic oxygen therapies and preserved more endothelial cells than conventional therapy.