We read with interest the article by Pires et al published in the October 2002 issue of the ARCHIVES.1 The authors found that the retinal thickness analyzer (RTA) detected statistically significant retinal thickening in 24 of 28 eyes with mild nonproliferative diabetic retinopathy and an absence of clinically manifest retinal edema. The magnitude of RTA-defined retinal thickening ranged from 0.3% to 73.5% more than the reference values. Ten patients were classified as having no clinically visible diabetic retinopathy, and the remaining 18 had minimal retinopathy. The authors reported that the RTA detected statistically significant retinal thickening in all 10 eyes with no clinically visible retinopathy. Interestingly, the optical coherence tomographer (OCT) only detected retinal thickening in 3 of 28 eyes. The authors concluded that the RTA is a more sensitive instrument for the detection of retinal thickening in patients with an absence of, or minimal, diabetic retinopathy.