The interpretation and diagnostic value of the ophthalmic artery pressure measurement and ocular pulse timing modes of the 500 mm Hg vacuum OPG-Gee ocular pneumoplethysmograph were evaluated in 65 patients who underwent aortic arch and cerebral arteriography. Data analysis revealed the OPG-Gee differential tracing which electronically compares and amplifies differences between the right and left eye pulse waveforms to be of little value. In predicting the presence of a greater than or equal to 50 percent diameter unilateral stenosis, an eye-eye pulse interval of greater than 15 msec was 82 percent accurate, a greater than or equal to 5 mm Hg ophthalmic artery pressure difference was 77 percent accurate and when combined these two criteria were 84 percent accurate. Neither of the criteria intended to detect bilateral carotid lesions, eye-ear pulse interval nor ophthalmic/brachial pressure index, were reliable. Ocular pulse timing was found to be highly specific but insensitive to hemodynamically significant carotid disease. OPG-Gee pressure determinations were more sensitive but lacked specificity. In combination, these criteria allowed identification of unilateral hemodynamically significant lesions with a sensitivity of 83 percent and a specificity of 86 percent. If used to detect more severe degrees of arteriographic stenosis, 60 and 70 percent diameter reduction, the overall diagnostic accuracy of these techniques was not improved. These results do not justify the use of the OPG-Gee instrument as a single noninvasive test for carotid arterial disease.