The value of duplex surveillance and the significance of contralateral carotid disease after endarterectomy have been assessed. Three hundred five patients were observed prospectively after carotid endarterectomy for a median time of 36 months (range, 6 to 96 months), with duplex surveillance performed at 1 day; 1 week; 3, 6, 9, and 12 months; and then each year after endarterectomy. Thirty patients (10%) had ipsilateral symptoms (13 strokes, 17 transient ischemic attacks [TIAs]) at a median time of 6 months (range, 0 to 60 months). Life table analysis demonstrated that ipsilateral stroke was equally common for patients who had > or = 50% restenosis (3% at 36 months) and those who did not (6% at 36 months, p > 0.5). Twenty-three patients (8%) developed symptoms (stroke 5, TIA 14) attributable to the contralateral carotid artery at a median time of 9 months (range, 0 to 36 months) after endarterectomy. By life table analysis, 40% of patients with 70% to 99%, 6% with 50% to 69%, 1% with < 50% contralateral internal carotid stenosis, and 5% with contralateral carotid occlusion at the time of endarterectomy had a contralateral TIA in the 36 months after endarterectomy (p < 0.01). However, contralateral stroke was not significantly more common for patients with severe contralateral internal carotid stenosis demonstrated at the time of endarterectomy (< 50% stenosis, 0%; 50% to 69%, 3%; 70% to 99%, 7%; occlusion, 6% stroke rate at 36 months). Seven of the 32 patients who developed progression of contralateral disease had a TIA, compared with 11 of 227 patients who did not develop progression of contralateral disease (p < 0.01). None of the 12 patients who progressed from a < 70% to a 70% to 99% contralateral stenosis had a stroke. After carotid endarterectomy restenosis is rarely associated with symptoms; contralateral stroke is rare and is not associated with progressive internal carotid artery disease suitable for endarterectomy. This study has shown no benefit from long-term duplex surveillance after carotid endarterectomy. Selective clinical follow-up of patients who have high-grade contralateral stenoses would appear more appropriate.