Age-related macular degeneration (AMD) is a neurodegenerative disease that causes adult-onset blindness. There are 2 forms of this progressive disease: wet and dry. Currently there is no cure for AMD, but several treatment options have started to emerge making early detection critical for therapeutic success. Analysis of the eyes of Abca4(-/-)Rdh8(-/-) mice that display light-induced retinal degeneration indicates that 11-cis-retinal and docosahexaenoic acid (DHA) levels were significantly decreased as compared with the eyes of control dark-adapted C57BL/6J mice. In addition, exposure to intense light correlated with higher levels of prostaglandin G2 in the eyes of Abca4(-/-)Rdh8(-/-) mice. Intense light exposure also lowered DHA levels in the eyes of wild-type C57BL/6J mice without discernible retinal degeneration. Analysis of human serum from patients with AMD recapitulated these dysregulated DHA levels and revealed dysregulation of arachidonic acid (AA) levels as well (∼32% increase in patients with AMD compared with average levels in healthy individuals). From these observations, we then built a statistical model that included levels of DHA and AA from human serum. This model had a 74% probability of correctly identifying patients with AMD from controls. Addition of a genetic analysis for one of the most prevalent amino acid substitutions in the age-related maculopathy susceptibility 2 gene linked to AMD, Ala(69)→Ser, did not improve the statistical model. Thus, we have characterized a reliable method with the potential to detect AMD without a genetic component, paving the way for a larger-scale clinical evaluation. Our studies on mouse models along with the analysis of human serum suggest that our small molecule-based model may serve as an effective tool to estimate the risk of developing AMD.