Many tests are used to examine visual fields by confrontation, but such methods have not been thoroughly compared with an accepted reference standard. The choice of test might affect the identification of subtle defects in the visual field. We prospectively compared seven confrontation field tests with full-threshold automated static perimetry among 138 outpatients in an eye clinic. Our primary outcome was detection of a defect in the visual field. With automated perimetry, most field defects were small or shallow. Most confrontation field tests were insensitive in the identification of field loss. The most sensitive method was examination of the central 20 degrees visual field with a small red target (73% [95% CI 63-82]). Assessment of the visual field should thus include such a test.