A randomized controlled trial was performed to evaluate patch angioplasty for patients undergoing carotid endarterectomy. There were 213 patient episodes affecting 148 men and 65 women, with 109 allocated to patch angioplasty. Following surgery six patients suffered transient ischaemic attacks but these did not delay discharge fi om hospital. Six individuals (four parched operations, two not patched) required re-exploration for postoperative haemorrhage and eight (two patched procedures, six not) had potentially serious neurological problems after operation. Of these eight patients, four (none receiving patch angioplasty) underwent re-exploration and in each case a clot was removed and a patch inserted; three of the four made a good long-term recovery. The other four patients suffered completed strokes from which one died. Two further patients (one patched procedure, one not) died after operation from myocardial events, giving an overall 30-day stroke or mortality rate of 2.8 per cent. Objective follow-up assessment,vith duplex scanning at I year was completed by 94.8 per cent of patients; significantly more vessel restenoses and occlusions were observed in those not receiving patches (P<0.01). Patch angioplasty reduces the number of immediate postoperative complications, and significantly lowers vessel restenosis and occlusion rates at 1 year after operation.