Automated perimetry has changed visual field testing considerably in recent years. What was considered an art has become an exercise in interpreting a set of data points obtained mechanically. Automated perimetry saves ophthalmologists time, which ideally should allow for more visual fields to be obtained on patients with unexplained vision loss. However, one must still keep in mind that automated perimetry still depends on the subjective responses from the patient. More important, automated perimetry has made interpretation of visual field defects, especially those due to occipital lesions, more difficult. For example, macular sparing may not be reflected, especially with programs limited to the central 24° or 30°. A 10° field may be required to show macular sparing. Also, sparing or involvement of the temporal crescent will not be shown with 24° or 30° visual fields. The limitation of most programs may lead to the appearance of incongruity when in fact the field is indeed congruous. Sometimes, a small homonymous hemianopic scotoma will be detected when one eye is tested but will be completely missed when the other eye is tested, giving the false impression that the visual loss is monocular. This is especially problematic if the patient also falsely interprets his or her homonymous loss of vision as monocular. Such individuals may complain of loss of vision in one eye when in fact it is one half of their visual field that is defective. The strategy of automated testing on either side the vertical and horizontal meridians may lead to the false impression that field defects respect the vertical or horizontal meridian when they do not. Automated perimetry should make it possible to test more patients with unexplained vision loss, but all automated visual fields must be interpreted with caution and, when necessary, substantiated with some other method, such as the tangent screen, which remains the most powerful method of detecting the size, shape, and density of visual field defects. Because most ophthalmologists no longer use tangent screen testing, at least an Amlser grid should be used to qualify the nature of a paracentral visual field defect.