Background: The aim of this study was to evaluate the relationship between structural and functional changes identified with the Rarebit fovea test (RFT) in diabetic patients over two years. In addition, we evaluated whether the RFT changes at baseline can predict vascular changes detectable by conventional screening methods four to six years later. Methods: Forty-two patients with diabetes and 42 age-matched healthy subjects underwent an initial examination. Two years later, 25 of the diabetic patients and 20 of the controls were re-examined in the same way. Four to six years later, 40 of the diabetic patients underwent a standard screening examination including fundus photography. After two years, all subjects were examined with the RFT and visual acuity (VA) was assessed. In addition, optical coherence tomography (OCT) and fundus photography were performed on the diabetic patients. At a screening examination of the diabetic patients after four to six years, fundus photographs were reviewed. Results: After two years, a marked difference in RFT results was observed between the 20 normal subjects and the 25 patients with diabetes. Results from other tests (VA and central retinal thickness, as measured with OCT) were unchanged compared with the initial examinations. Fundus photography four to six years later of 40 of the 42 diabetic patients showed an increased incidence of retinopathy unrelated to the RFT findings at baseline. Conclusions: The findings in the current study indicate that the RFT might detect macular dysfunction in diabetic eyes without microvascular changes. This dysfunction increased during a two-year period and presumably reflects neural impairment in diabetes mellitus but did not predict development of retinopathy during the four to six year period.