The authors present results of a study in which certified visual acuity examiners in the Macular Photocoagulation Study Group, performed independent replicate refractions and visual acuity measurements on both eyes of patients whose visual acuities ranged from 20/20 to less than 20/800. Patients with neovascular age-related macular degeneration (AMD), ocular histoplasmosis, or idiopathic neovascularization were represented. A total of 328 observations for 164 eyes from 82 patients were available for analysis. Overall reliabilities of the refraction data and the visual acuity data were 98.9% and 98.7%, respectively. Differences between replicate measurements were small; 87% were less than one line. However, differences were greater in patients with visual acuity less than 20/100 and in patients with AMD. These results indicate that the reliability of refraction and visual acuity measurements under a standard protocol is high and may depend on both visual acuity level and disease process. If a standard, systemic procedure is used for refraction and visual acuity testing, the clinician may be able to differentiate between true changes and measurement error, even among low vision patients with AMD.