OBJECTIVES: To assess the cost-effectiveness of various screening intervals using indirect ophthalmoscopy performed by ophthalmologists for detecting diabetic retinopathy among type-2 diabetic patients from hospital and societal perspectives in Thailand. METHODS: A Markov model of diabetes retinopathy was constructed based on Eastman model. A cohort of 10,000 newly diagnosed, type-2 diabetic patients age 40 years was simulated until the age of 75 years or death. Transition probabilities were obtained from literature review, while unit costs and utilization patterns in Thai population were derived from a hospital and expert opinions. Costs of screening and treatment were calculated using a microcosting technique. Incremental cost-effectiveness analyses were performed based on a cohort simulation and presented as incremental Baht per blindness prevented. A series of sensitivity analyses were performed. RESULTS: In base-case analysis using a hospital perspective, an incremental cost-effectiveness ratio (ICER) of screening every four-years compared to no screening was 85,976 Baht per blindness prevented. Comparing to the next preceding screening frequency, the ICER of screening frequency every four-years, every three-years, and two-years was 62,806, 70,554, 95,865 Baht per blindness prevented, respectively. When varying the cost of screening and laser treatment, probability of seeking medical treatment among unscreened, probability of being treated with vitrectomy, and annual mortality rate, the results remained similar to the base-case analysis. However, when we performed analysis using a societal perspective, all screening patterns resulted in cost-saving. CONCLUSIONS: Our findings indicated that an incremental cost incurred from increasing frequency in all screening intervals was less than 100,000 Baht or US$2,500 per one blindness prevented. Annual diabetic retinopathy screening seems to be a cost-effectiveness intervention, especially from a societal perspective. Our findings can be a very useful information to aid healthcare policy decision makers during their decision making process.