Intensive glucose therapy can protect the retina of individuals with diabetes, but it is unknown if it provides the same protection to patients with different severity of diabetic retinopathy (DR). We finally included DR-related studies involving intensive glucose control with large sample size and long follow-up time, including five large and high-quality randomized clinical trials (RCTs): DCCT, UKPDS, ACCORD, AdRem, and VADT. With DCCT as a reference, we supposed a DR severity threshold that is verified by other RCTs then. We found that individuals who have DR lesions that are equivalent to or less severe than moderate NPDR achieve benefits for the retina by intensive glycemic control. However, these are realized only if the HbA1c in type 1 or type 2 diabetic patients is reduced at least by 0.8% versus the control group or it is reduced to 3 years of intensive glucose control is required. If the severity of DR lesions is worse than moderate NPDR, intensive glycemic control may not bring benefits.