A total of 1780 patients without symptoms of cerebral ischemia undergoing coronary artery bypass grafting (CABG) were screened before surgery for carotid stenosis by pneumophlethysmography. An abnormal test was defined as a difference in ophthalmic artery pressures of > or = 5 mmHg or ophthalmic-brachial pressure index < or = 0.69. Some 99 patients (5.6%) had an abnormal ocular pneumoplethysmographic measurement (89 unilateral, ten bilateral). Of these, 26 patients underwent prophylactic carotid endarterectomy before CABG (group 1), while the remaining 73 patients had reconstruction without previous carotid endarterectomy (group 2). A total of 100 patients (group 3) with normal ocular pneumoplethysmographic results were used as controls. The three groups were comparable with respect to age, diabetes, hypertension, smoking and severity of coronary artery disease. Early (30-day) postoperative stroke rates were 0 and 4% (n = 3) for groups 1 and 2 respectively, and 0% for group 3 (P > 0.07). Early mortality rates after CABG for groups 1 and 2 were 4% (n = 1) and 1% (n = 1), respectively and 2% (n = 2) for groups 3 (P > 0.4). Late follow-up (mean 48 months) demonstrated stroke rates of 0% for group 1, and 10% and 4% for groups 2 and 3 (P > 0.08). The early mortality and stroke rates in the ten patients with bilateral abnormal ocular pneumoplethysmographic values were 0 and 0%. However, late strokes occurred significantly more often (43%) in patients with bilateral abnormal results compared with those with unilateral abnormal findings when both groups did not undergo carotid endarterectomy (P < 0.04).(ABSTRACT TRUNCATED AT 250 WORDS)