The value of the information obtained from Goldmann manual kinetic perimetry beyond 30 degrees was examined. Of 229 randomly selected patients in a University eye clinic who had visual fields performed for reasons other than glaucoma or ocular hypertension only 3 patients had abnormalities confined to the PVF of one or both eyes. In none of these three patients was the PVF necessary to detect disease (Graves' disease, 2 cases; retinoschisis, 1 case). The PVF was useful in determining the localization of the disorder and/or the therapeutic management in 14 patients of whom 4 of these had retinitis pigmentosa and 5 had other disorders where the PVF showed the extent of the retinal damage. For ergo-ophthalmologic purposes the PVF was useful in 45 patients; most frequently because the extent of abnormality provided a basis for warning the patient. In some cases the PVF was considered to be useful for economic disability determination or to exclude significant PVF defects in a patient with only one visually useful eye. In 77 patients the PVF of each eye was abnormal but not of ergo-ophthalmologic significance. If these data can be extrapolated to automated static perimetry, there will be a very great incremental cost for any clinically useful information obtained from the examination of the PVF. Because the cost-effectiveness of the examination must be compared with competing methods of obtaining information, it is proposed that the PVF be examined (1) whenever indicated for ergo-ophthalmologic reasons, or (2) when the CVF examination does not resolve a clinical problem for which there is a reasonably high probability that (a) additional clinically useful information will be obtained by examination of the PVF after the results of the CVF examination have been analyzed, or (b) the eye is likely to have a condition that can be detected or followed best by PVF examination.