Automated perimetry was performed on both eyes of 54 normal subjects, 36 patients with ocular hypertension and normal visual fields, and 20 patients with early glaucomatous visual field loss to evaluate false-positive errors, false-negative errors, fixation losses, consistency of double determinations, and testing time. For all subject groups and response measures, large interindividual variation was found. No meaningful age-related changes were obtained for false-negative errors, false-positive errors, fixation losses, or consistency of double determinations. Contrary to earlier reports, we found a low number of normal subjects and patients exceeding the 33% false-positive and false-negative limits established for the Humphrey Field Analyzer. A large number of normal subjects and patients, however, exceeded the 20% limits for fixation losses.